DEPARTMENT OF POLITICAL SCIENCE Master’s Thesis: 30 higher education credits Programme: Master’s Programme in Political Science Date: 02.01.2018 Supervisor: Elin Naurin Words: 18115 UNDER PRESSURE: DEMOCRATIC MANDATES AND POLICY OUTCOMES IN TURBULENT TIMES. EVIDENCE FROM GREECE Natalia Natsika i Abstract To what extent do political parties under conditions of extraordinary pressure manage to fulfill the promises they made at elections? While systematic evidence from an international context points to parties under ‘stable’ times fulfilling high levels of election pledges (Thomson et el., 2017), research and knowledge in contexts of crisis is lacking. This study offers both the first empirical investigation of Southern European governing parties’ policymaking ability during the Great Recession, and the first examination of pledge fulfillment in Greece the past twenty years. In this project, I examine the strength of the program-to-policy linkage in Greece under the recent period of the 2008 economic crisis, with a parallel aim to assess the applicability of conclusions raised in Thomson et el. (2017) under conditions of pressure. Using PASOK’s (2009) and New Democracy’s (ND) (June 2012) election programs, I create a unique dataset consisting of 652 pledges and covering two consecutive electoral cycles. I then evaluate the fulfillment of 120 pledges made by PASOK and ND prior to the formation of the 2009-2012 and 2012-2015 governments accordingly. The findings indicate that PASOK fulfilled a substantially high level of its most important promises, while ND was unable to follow through on a vast majority of its pledges. Moreover, this study provides unique results for the impediments of maintaining the status quo under crises, and highlights the conclusive importance of institutional constraints on policymaking. Although explorative, this project covers an important gap in the literature, and offers insights on various tribulations of governance. Keywords: representative democracies; crises; economic crises; pledge fulfilment; election promises; Greece; Great Recession. ii Abbreviations DIMAR Democratic Left EC European Commission ECB European Central Bank EU European Union GDP Gross Domestic Product IMF International Monetary Fund KKE Communist party of Greece LAOS Popular Orthodox Rally MoU(s) Memorandum(s) of Understanding ND New Democracy PASOK Panhellenic Social Movement SYRIZA Coalition of Radical Left Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece iii Table of Contents 1. Introduction.............................................................................................................................1 2. Literature Review....................................................................................................................4 2.1 Representative Democracies under Crises.......................................................4 2.2 Representative Democracies under the Great Recession..................................7 2.3 Pledge Fulfillment Theory and Practice.........................................................10 2.4 Theoretical Approach ………………............................................................12 3. Research Design....................................................................................................................13 3.1 Case: Elections/Governments........................................................................14 3.2. Hypotheses…………………………………………………………………15 3.3. Method……………………………………………………………………..18 3.3.1. Definition of Pledges………………………………………20 3.3.2. Coding of Pledges…………………………………………21 3.3.3. Evaluation of Fulfillment…………………...……………..22 3.3.4. Comparative Analysis and Limitations……………………23 3.4. Selection of Promises and Sources for Fulfillment…………………………23 4. Results……………………………………………………………………………………...25 4.1. Pledge Making in Greece…………………………………………………..25 4.2. Pledge Fulfillment in Greece……………………………………………….29 5. Discussion………………………………………………………………………………….33 5.1. Limitations…………………………………………………………………36 6. Conclusion……………………………………………………………………………….....37 7. References………………………………………………………………………………….39 Appendix…………………………………………………………………………...…………46 Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece iv List of Tables Table 1. Main political events in Greece: 2009-2015…………………………………………14 Table 2. Total number of pledges made in PASOK’s and ND’s election programs per category.………………………………………………………………………………...….…25 Table 3. Number of pledges analyzed for fulfillment per party and category.……………….27 Table 4. Fulfillment of pledges in Greece…………………………………………………….29 Table 5. Pledge fulfillment by PASOK and ND per status quo/change ………………………31 Table 6. Fulfillment of ND’s pledges: negotiation with the ‘Troika’…………………………32 Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 1 1. Introduction To what extent do political parties during periods of pressure manage to fulfill their campaign promises and enact policies closely aligned to their electoral commitments? The literature has long established that the onset of a crisis may lead to the destabilization of the ‘normal’ functioning of representative democracies, that could then affect and disturb incumbent parties’ representation functions (Kriesi, 2015; Baumgartner & Jones, 1993; 2002). In the short-run, these functions may be affected because political parties may proceed to an across-the-board political consensus (e.g. Lijphart, 1996; Chowanietz, 2011), may face a drastic reduction of their policymaking space and their room to maneuver (Hellwig, 2001; Kosmidis, 2014), or might be forced to abandon previous political programs (Stokes, 2001). In the long-run, the structural party system changes caused by large-scale crises may lead to a crisis of political representation (Mainwaring, Bejarano & Lengómez, 2006), that changes the external conditions of government, and might further disrupt the ‘normal’ functioning of representative governments (Kriesi, 2015. While systematic evidence is lacking, we know very little on the extent to which conditions of generalized pressure further affect governments’ policymaking1 ability. In the aftermath of the Great Recession2, Southern European democracies experienced increased external and domestic strains. The deep economic recession, the harsh austerity measures, the external conditionality3, and in many cases the crisis of domestic politics, greatly challenged the then governing parties. These pressures restricted national governments’ maneuvering space across different countries, and affected their ability to respond to, and represent their constituencies (Freire, Lisi, Andreadis & Viegas, 2014; Moury & Freire, 2013). In this direction, it has been a commonly shared view that many Southern European governments reneged on their election promises to meet the new economic demands (Bosco & Verney, 2012; Moury & Freire, 2013; Kriesi, 2015). However, no research has yet examined this assumption in the context of the 2008 crisis. Parties’ ability to fulfill election promises in times of crisis is a particularly interesting and politically relevant question, as it offers the opportunity to test the normative 1 In this thesis, ‘policymaking ability’ is used as equivalent to the ability to fulfill election pledges. The term ‘policymaking capacity’ is used for more general forms of policymaking. 2 The Great Recession is the international economic crisis that hit Europe in late 2007, named after the Great Depression of the 1930s. It started with the breakdown of the construction and property booms in Spain and Ireland, and affected almost all European states the following years (Stiglitz, 2010). The terms ‘Great Recession’, ‘Eurozone crisis’, ‘2008 crisis’, or ‘recent crisis’ are used interchangeably. 3 ‘External conditionality’, ‘international supervision/surveillance’: the terms refer to EU agencies and the IMF placing institutional, financial, and policy constraints to national governments. Terms are used interchangeably. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 2 expectations, as well as the bearing of empirical findings of democratic theories under destabilizing conditions. From a normative lens, the traditional democratic models of the ‘mandate theory of democracy’ and the ‘responsible party principle’ emphasize as a core feature of well-functioning democracies the ability of political parties to introduce policy priorities into election manifestos to voters, and to deliver on their pre-electoral commitments upon taking office (APSA, 1950; Downs, 1957; Klingemann, Hofferbert & Budge, 1994; Pierce, 1999). Empirically, a recent comparative piece by Thomson et al. (2017) investigated more than 20000 pledges from political parties in 12 countries across time, and found that parties in ‘stable’ Western democracies fulfill a majority of their election promises. Yet, and while economic conditions are systematically found to affect policy outcomes (Thomson et al., 2017), pledge fulfillment under large-scale economic crises remains an under-researched field. Combining the above, the aim of this thesis is to determine the strength of the program-to-policy linkage, i.e. the congruence between parties’ election manifestos and their subsequent policy outcomes, in a democratic setting under significant pressures. The analysis examines a critical case -Greece-, under a turbulent period -from 2009 to 2015- to assess the applicability of the conclusions lifted forward by Thomson et al. (2017) in a democracy in crisis. Does a political system under pressure achieve the normative expectations of the mandate theory to the same extent as ‘stable’ democracies? Do governing parties under a deep and prolonged economic recession manage to fulfill their election promises to the same extent as parties under ‘normal’ times? These are the general questions that drive this thesis, and the analysis attempts to offer insights on potential tribulations of governing in turbulent times Greece represents a characteristic case of a democratic system under extraordinary pressure. During the Eurozone crisis, the country experienced the highest debt- to-GDP levels, the deepest economic recession, the strictest austerity programs, the longest period of external supervision, and the most turbulent political developments than any other European country (e.g. Freire et al., 2014; Kriesi, 2015). The debt crisis necessitated the request of financial rescue loans from the European Commission (EC), the International Monetary Fund (IMF), and the European Central Bank (ECB) (i.e. the ‘Troika’). The bailout funds were accompanied by two4 Memoranda of Understanding (MoUs) under the period of examination, i.e. loan agreements that conditioned harsh austerity measures and imposed policy constraints on governments. As a result, the country lost a quarter of its economic output, social inequality increased significantly, and unemployment reached 25 percent. In addition, an equally deep 4 3 after August 2015 Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 3 political crisis followed, characterized by party system fragmentation, political polarization, and a crisis of governance (Sotiropoulos, 2012; Vasilopoulou & Halikiopoulou, 2013; Tsirbas, 2015). How likely are the above pressures to affect Greek parties’ ability to act on their pre-electoral appeals? Put differently, to what extent are the conclusions raised in an international context confirmed in Southern European democracies under the recent crisis, and particularly in crisis-hit Greece? Taking Thomson et al. (2017) into account, but also previous findings from Greece (Kalogeropoulou, 1989) that reported high fulfillment rates for PASOK’s 1981 election promises, I examine the extent to which the economic recession, the austerity policies, the international surveillance, and the political implications of the crisis may lead to lower levels of pledge fulfillment for the governments of PASOK (2009-2012) and the New Democracy (ND) coalition (2012-2015). To evaluate this, I first conduct a content analysis on PASOK’s and ND’s election programs to identify and categorize pledges which results in a unique dataset of 652 pledges covering two consecutive electoral cycles. Then, I assess official legislative enactments to test fulfillment, and finally I compare my findings to the study of Thomson et al. (2017). This study makes three main contributions. First, it adds to the broader literature on representative democracies under crises. A series of studies have showed that under a war (Lijphart, 1996; Riker, 1964; Bueno de Mesquita, 1981), or a terrorist attack (Chowanietz, 2011; Indridason, 2008) parties might ‘suspend’ their representation role and come together in coalitions. Economic crises have also been suggested to disrupt incumbents’ functions by making them undermine their ‘responsive’ at the cost of their ‘responsible’ role (e.g. Mair, 2009). By examining parties’ ability to fulfill election pledges in times of crisis, I add further insights on parties’ policymaking capacity and ability under conditions of pressure. Secondly, the analysis contributes to the pledge fulfillment research in two important ways: first, it adds a case of a democracy under pressure, that furthers the understanding on the strength of the program-to-policy linkage in unstable contexts. Second, it enriches the under-researched field of pledge fulfillment in Greece, where the only existing study is by Kalogeropoulou (1989) on the fulfillment PASOK’s 1981 election pledges. Finally, this work contributes to the literature on the European experiences under the Great Recession. Most of the existing studies focus on voters’ behavior and parties’ electoral fortunes (Kriesi, 2012; 2014; Bartels, 2014; Bartels & Bermeo, 2014; Hernandez & Kriesi, 2016), or governments’ responses to the crisis (e.g. Bermeo & Pontusson, 2012), but no study examines governing parties’ ability to fulfill their electoral commitments. I add to the Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 4 extant literature the first empirical investigation of pledge fulfillment by Southern European governments during the Eurozone crisis. The remainder of this thesis is organized as follows. In the second chapter I introduce previous research on representative democracies under crises, pledge fulfillment theory and practice, and present the theoretical approach. The third part includes the research design, where I describe the case, formulate my hypotheses, and present my methodology. Moving on, the fourth chapter presents the results, starting with the making of Greek pledges, and followed by the analysis of their fulfillment. The fifth section offers a discussion of the overall findings, and addresses the limitations of this project, and finally, chapter six concludes by summarizing the main findings and drawing suggestions for future research. 2. Literature Review This overview includes previous research on representative democracies under crises, different evidence from European countries during the Great Recession, and pledge fulfillment research. Given the variety of the processes in the chain of democratic representation, and the equivalent potential disruptions under crises, the scope of the first two parts of this overview is not to be exhaustive. The aim is rather to identify the mechanisms and the factors that pressurize, and potentially affect the ability of governing parties to fulfill their pledges. While I mainly focus on economic crises, since arguably they shape the context of this thesis, I also attach crisis outcomes from different strands of literature that incorporate the analysis into a broader perspective of policy enactment in crisis situations. 2.1. Representative Democracies under Crises The study of democracies under crises is highly relevant and crucial for political science theory and practice, as it can provide a deeper understanding of the different trajectories democratic processes can take under conditions of pressure. Crises are critical events or ‘tipping points’ (Gladwell, 2001), in that they have the potential to cause important changes in the wider political system in which they occur (Walby, 2015). The emergence of a crisis, along with the increased uncertainty and the potential adverse outcomes, may lead to the destabilization of the institutional, political and economic features that facilitate the ‘normal’ functioning of representative democracies (Kriesi, 2015). Under extraordinary conditions, incremental policymaking may no longer be applicable, and the ‘equilibrium’ of ‘normal’ politics is likely to be interrupted (Baumgartner & Jones, 1993; 2002; Kriesi, 2015). Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 5 At the onset of a severe crisis, incumbent parties have the responsibility of coping with the crisis, and during this process, they might face increased challenges that can greatly affect and disturb their representative functions. Since research on governing parties’ policymaking capacity and ability under crises is scarce, below I discuss ways in which directly or indirectly their functions are disrupted, and their ability to fulfill election pledges might be hindered. First, literature has long stressed that the outbreak of a crisis may lead to an across- the-board political consensus. Lijphart (1996, p. 263) has suggested that the existence of an external threat enables power-sharing consensus among parties; in a similar vein, Riker (1964) and Bueno de Mesquita (1981) have argued that in the face of a war, political leaders might collaborate or converge in ‘unity’. Similar effects have been reported for events like terrorist attacks (see Chowanietz, 2011; Indridason, 2008), under which policy makers may show greater agreement than under ‘normal’ times, and put party and ideological conflicts aside (Weschle, 2017). Two possible implications of these suggestions could be extracted in relation to my project; first, that when governing parties under crises form ‘grand’ coalitions, - ‘pausing’ therefore their role of responding to the preferences of the electorate-, may also ‘pause’ their role regarding the fulfillment of their promises. Secondly, and in contrast, when parties collaborate in consensus and put any disagreements aside, might be more likely to fulfill their promises. Secondly, a crisis may significantly restrict governments’ policymaking and maneuver space (Hellwig, 2001; Kriesi, 2015). The room to maneuver refers to the extent to which governments can freely suggest, draft and implement policies (Hellwig, 2001). Governing parties’ policy discretion is considered to have been limited the past three decades due to globalization and Europeanization processes, and the various constraints that stem from membership in supranational institutions (e.g. Hays 2003; Lobo & Lewis-Beck, 2012). In an extraordinary situation, extra layers of constraints might be added, because of the initiatives required (Kosmidis, 2014). For example, in an international economic crisis, the global nature of the economy shifts the responsibility of dealing with the crisis from the national to the international level (Armingeon & Baccaro, p. 161-2). Especially in cases of state bankruptcy, national governments are often compelled to resort to international institutions for financial support, which is highly likely to prevent them from designing and implementing policies (Kosmidis, 2014; Kriesi, 2015). This consideration is of particular relevance to my study, since the context under examination greatly resembles the one described here, both in economic terms, but also in regard to the different sets of political actors involved. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 6 Thirdly, a critical event may cancel out political leaders’ previous programs and cause ‘policy switches’ (Stokes, 2001). The urgency of the situation calls for extraordinary measures that might contradict parties’ political decisions, or exceed previous budget calculations. Governments may be forced to abandon their campaign promises and implement policies that are not in line with their manifestos or their ideological orientations (Stokes, 2001). A characteristic example is the Latin American experience of the 1980s and 1990s, when many left-wing governments, during a deep economic crisis, had to change course and adopt neoliberal reforms under the pressure of international institutions (Stokes, 2001; Kriesi, 2015). The example from Latin America reminds in many regards the experience of Southern European democracies, and the ‘policy switches’ could be the potential eventual outcome in my investigation. Prolonged and deep in nature crises, such as large-scale international economic crises, apart from the challenges they cause on incumbent parties in the short-run, they can further produce more consequential and long-term effects on representative democracies, that can indirectly affect parties’ policymaking ability. During international economic crises, systems of production in a global scale go through severe breakdowns and transformations; the economy recesses, growth stagnates, unemployment rises, and social inequality grows. When these processes interact with the political system, they cause or trigger restructurings and alternations in the distribution of power among political parties and social groups in an abrupt way (Polanyi, 1957). Large-scale international economic crises are viewed as critical junctures, in that they can lead to a crisis of democratic representation, or a political crisis, characterized by electoral de- and realignments, new sociopolitical formations and dynamics, and institutional change (Bermeo & Pontusson, 2012, p. 27; Mainwaring, Bejarano & Lengómez, 2006). According to Mainwaring et al. (2006), a crisis of representation pertains to the co-existence of increased levels of popular distrust and electoral volatility, the collapse of the party system, decreasing voter turnout and increasing support for ‘outsiders’. Combined, these procedures may trigger the decline of traditional cleavages and of established parties, with the parallel emergence of new social divides and new political challengers (Kriesi, 2015). In Latin America for example, the market liberalization that was undertaken in response to the 1980s crisis, created programmatic de-alignment of the party competition, and party system destabilization (Remmer, 1991; Roberts, 2013; Lupu, 2013). Such radical changes are related to political parties’ governing functions, and therefore might be relevant for my project. A destabilized party system may produce weaker and more unstable governments, disturb the connection Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 7 between electoral outcomes and policy outputs, and eventually facilitate policy discontinuations and mandate unresponsiveness5 (Stokes, 2001; Ryu, 2009; Kriesi, 2015). To sum up this review, literature suggests that political parties under crises experience a series of pressures that disturb, limit, or pause their ability to govern or represent. In addition, the potential destabilization and weakening of the party system, may also affect governments’ functions. In the following section, I examine how these recommendations apply in European democracies during the Great Recession. 2.2. Representative Democracies under the Great Recession The consequences of the Great Recession that hit Europe in 2008 overall confirm the suggestions lifted forward in the literature regarding the impact of crises on representative democracies. However, the consequences were borne unequally among different states; while almost all European democracies were affected to some extent, the types of pressures and the size of the effects were shared disproportionately among the European North and South (Bermeo & Pontusson, 2012; Bartels & Bermeo, 2014). This unequal bearing of the crisis is contingent -in broad terms- on the levels of institutionalization of party systems, on the type of economic crisis, and on the different responses to the crisis (Kriesi, 2015). Northern European democracies overall possess well-institutionalized party systems and have experienced long periods of economic prosperity and stability. These characteristics provided them with a more resilient environment, and resulted in a relatively quick economic recovery and moderate overall political effects (Bermeo & Pontusson, 2012; Bartels & Bermeo, 2014). The political repercussions of the crisis mainly included the erosion of mainstream parties’ representation functions, their electoral decline with a parallel emergence of new parties, and the partial transformation of the party system (Kriesi, 2015). More specifically, the 2008 crisis is considered to have accelerated a decades-long party system change. According to Mair (2006; 2009), mainstream parties in advanced democracies, due to changes in the environment of party competition and the demanding conditions of multi-level governance, have strengthened their governing role at the cost of their representative function. The crisis is perceived to have accelerated this divide, and this has resulted in a division between parties that ‘govern’, and parties that ‘represent’ (Kriesi, 2015). The former, i.e. mainstream parties, experienced important losses, while the latter, i.e. (right-wing) populist parties, saw a sharp rise in their popular support (Bartels, 2014; Kriesi, 2012). 5 Mandate responsiveness (or unresponsiveness) is another term for pledge fulfillment. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 8 In contrast, the economic and political consequences of the crisis in Southern Europe were larger and more consequential. Southern European countries generally possess weaker party systems, poorer management of public resources, and larger sovereign debts (Cunha, 2008). The onset of the economic crisis stepped on, and brought to light, these institutional deficiencies, which, along with the nature of the crisis6, and the different responses, made the situation harder to manage (De Sousa, Magalhães & Amaral, 2014; Kriesi, 2015). The mounting sovereign debt of Southern European states necessitated their national governments to request loans from supranational institutions to avoid bankruptcy. As a prerequisite for receiving the funding, an international supervisory body consisting of EU agencies and the IMF, or the ‘Troika’, placed policy constraints on governments, that drastically reduced their maneuvering space, and greatly prevented them from having the control over policymaking (Armingeon & Baccaro, 2012, p. 161-2; Kosmidis, 2014; Freire et al., 2014). The policy constraints mainly pertained to the imposition of harsh austerity measures and reforms, which generally consist of cuts in public spending and increases in taxes. These policies had severe consequences on the economy, the society, and the political system. In the economic domain, austerity caused a deeper and longer economic recession, led to skyrocketing levels of unemployment, and to an increased social inequality (Cameron, 2012, p. 124). In the political sphere, austerity is linked to ruling parties’ electoral losses (Kriesi, 2014), political instability and social unrest (Blyth, 2013; Rüdig & Karyotis, 2014) and to changes in patterns of political representation (Freire et al., 2014; Teperoglou, Freire, Andreadis & Viegas, 2014). In relation to the latter, both the imposed austerity and the presence of the ‘Troika’, significantly affected the way (governing) Southern European parties represent their constituents. Mair (2011), following on his previous argument (2006), suggested that the debt crisis and the external intervention made political actors in the bailed-out countries accountable not only to their voters at the national level, but also to their international partners. In fact, studies find that the issue congruence between governments and their constituents increased importantly, as incumbent parties undermined their representative role and assumed a more ‘responsible’ stand toward governing (Moury & Freire, 2013; Freire et al., 2014; Teperoglou, et al., 2014). In addition, and in light of the above argument, a commonly shared perceived effect of the crisis is that Southern European governments abandoned their election promises 6 In most cases the crisis in Southern Europe pertained to, or evolved into a sovereign debt crisis, although there are important differences among countries (e.g. see Freire et al., 2014 for Portugal and Greece). Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 9 to meet the new demands (Bosco & Verney, 2012; Kriesi, 2015). No study however has examined further this view. Last, the economic hardships and the external pressures created a turbulent political climate, and in many cases caused a political crisis. On the supply-side of politics, the crisis led to a restructuring of the party environment through the decline of mainstream parties and the emergence of new political formations. These dynamics altered or weakened the composition of the cabinets, triggered early elections, and changed the nature of the party competition (Hernandez & Kriesi, 2016). In some extreme cases, they further caused a crisis of domestic politics, with the most characteristic example being that of Greece (Kriesi, 2015). Greece has been the country mostly hit by the crisis both in economic and in political terms. At the outbreak of the Great Recession, the country entered a never-ending phase of increased economic uncertainty, and prolonged recession and international surveillance (e.g. Featherstone, 2015; Tsirbas, 2015). The sovereign debt crisis led the Greek Prime Minister George Papandreou in April 2010 to request the first rescue loan from international and financial institutions in order for the country to serve its debt obligations. The requirements for receiving the funding, as conditioned by the MoUs, imposed harsh austerity measures such as large public expenditure cuts, increases in taxes, drastic cuts in public sector wages and pensions, the reduction of the size of the public sector, and the privatization of public assets (European Commission, 2013). As a result, the economy shrunk by 25 percent and unemployment increased alarmingly (Featherstone, 2015; Tsirbas, 2015). The economic grievances, along with the pre-existence of a low quality of government, led to the destabilization and collapse of the strong, stable, and long-existing, ‘two- party’ system (Sartori, 1976, p. 44; Pappas, 2003; Sotiropoulos, 2012), which was until then dominated by the socialist PASOK and the conservative ND alternating in power (Nicolacopoulos, 2005). In the 2012 elections, PASOK and ND were extraordinarily punished, and together lost around 45 percent of their vote share, while new political parties emerged (Kriesi, 2015). In addition, the agreement or disagreement with the MoUs and the austerity policies triggered the decline of the traditional ideological left–right cleavage, and created a new pro/anti MoUs and pro/anti EU and Eurozone split (Vasilopoulou & Halikiopoulou, 2013; Gemenis & Nezi, 2015). These developments led to an unprecedented crisis of governance and a fragmented party system and ideological spectrum, characterized by a profound inability of cabinet formation, coalitions instead of single-party governments, frequent cabinet reshuffles, exit of coalition partners, and earlier termination of governments’ terms (Tsirbas, 2015; Sotiropoulos, 2012; Gemenis & Nezi, 2015). Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 10 To sum up, this review examined how the Great Recession affected European representative democracies. Southern European countries were arguably more affected, as the deep recession, the harsh austerity, the extra constraints, and the changes in the party environment limited their incumbents’ ability to govern or respond to their voters. What has not been addressed yet is whether these conditions allowed incumbents to implement their election promises. Parties’ ability to fulfill election pledges during turbulent times is a particularly interesting issue, especially in the light of research findings that point to democratic governments following through on a majority of their promises (see Thomson et al., 2017). Below I describe this research field further. 2.3. Pledge Fulfillment Theory and Practice The fulfillment of parties’ election pledges is highly relevant and central to the theory and practice of representative democracy. Traditional democratic models like the ‘mandate theory of democracy’ and the ‘responsible party principle’ offer the core scheme under which the contemporary democratic political processes unfold (APSA, 1950; Downs, 1957; Klingemann et al., 1994; Pierce, 1999). These democratic theories put particular emphasis on the congruence between political parties’ electoral pronouncements and their subsequent government enactments (Mansergh & Thomson, 2007). The core assumption is that programmatic parties in competitive elections introduce their policy priorities to voters in cohesive and concise electoral manifestos to secure electoral support, and then carry out their programs upon taking office (Downs, 1957; Klingemann et al., 1994). Mansbridge (2003, p. 515) finds ‘the idea that during campaigns representatives made promises to constituents, which they then kept or failed to keep’ to be the core component of traditional models of representation. According to this so-called ‘promissory representation’, elected officials have a normative obligation toward their constituents to implement the electoral programs they put forward during election campaigns when they enter office (Mansbridge, 2003). Political parties and their manifestos are crucial in the representation process: parties are the main collective actors that link the electorate to the government, and are catalytic to political decision-making and implementation (Klingemann et al., 1994). Party manifestos are the means through which parties channel their positions to voters and the basic source of future governmental actions. If parties fail to enact policies aligned with their pre-electoral appeals, the core democratic function and linkage is arguably disturbed (Klingemann et al., 1994). On the other hand, if the congruence between election programs and policy outcomes Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 11 is strong, one of the most important indicators of a well-functioning democracy is achieved (Mansergh & Thomson, 2007). The empirical investigation of the fulfillment of electoral promises has largely confirmed the normative assumptions of traditional democratic theories. The pledge approach has a long and established tradition in political science research, and numerous studies on several countries show that on an aggregate level, parties manage to succeed a relatively and sometimes substantially high percentage of (at least partial) fulfillment of election promises, that ranges from 50 to 90 per cent (for example Pomper, 1968; Pomper & Lederman, 1980; Rose, 1980; Rallings, 1987; Kalogeropoulou, 1989; Royed, 1996; Thomson, 2001; Artés & Bustos, 2008; Naurin, 2011; 2014; Artés, 2013; Kostadinova, 2013). Thomson’s et al. (2017) study advances the pledge approach significantly by offering the first comparative analysis of pledge fulfillment in 12 countries. Their data derive from a comparative project with common definitions and reliability tests that allow for comparisons among different contexts. The 12 researchers primarily focus on the impact of power-sharing arrangements on pledge fulfillment, and their main findings indicate that parties in single-party governments, with and without legislative majorities, are more likely to fulfill their promises than parties in coalitions. Moreover, the scholars control for the impact of institutional, economic and time constraints on pledge fulfillment, and find a significant effect of economic and time resources, as well as the existence of institutional constraints, on pledge fulfillment. My project follows the same theoretical angle to assess the bearing of these conclusions on a ‘crisis’ context. The countries in Thomson et al. (2017) are advanced Western democracies under overall stable conditions. The only cases where some forms of uncertainties existed are Bulgaria, Ireland and Portugal. Bulgaria experienced increased political instability and economic transformations (Kostadinova, 2013), while Ireland and Portugal faced important economic hardships and received financial support from international institutions (Murphy, 2011; De Sousa et al., 2014). Yet, these challenges did not prove consequential for their governments’ ability to fulfill elections promises, and high levels of pledge fulfillment were observed. Overall, mandate responsiveness in times of crisis is an under-researched field. Crisis events or situations have not been properly examined (Naurin, 2009, p. 69), and only the work of Stokes (2001)7 approaches pledge fulfillment from a large-scale economic crisis 7 Stokes (2001) follows a somewhat different approach; she examines the reasons why Latin American governments abandoned their promises. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 12 perspective. While adverse economic conditions have been accounted for and found to matter for pledge fulfillment variation (see Mansergh & Thomson, 2007; Thomson et al., 2017), the extent to which large-scale crises condition or limit pledge fulfillment ability has not been sufficiently assessed. This study attempts to cover some important gaps in the literature. From the perspective of representative democracies under crises and the specific experiences of Southern Europe during the 2008 crisis, there is very little we know on governing parties’ policymaking ability. From a pledge fulfillment angle, both Thomson’s et al. (2017), and Kalogeropoulou’s (1989) conclusions are drawn from ‘stable’ contexts and periods, and research is lacking on different settings. In this project, I combine the above strands of literature and aim to examine the strength of the program-to-policy linkage in Greece, as well as to assess the extent to which the conclusions by Thomson et al. (2017), and by Kalogeropoulou (1989) can be applied and confirmed in a political system that does not meet these ‘stable’ contextual conditions. Below I present the theoretical approach, where I identify four aspects of ‘pressure’ that could potentially affect Greek pledge fulfillment. 2.4. Theoretical Approach Drawing on both external and domestic pressures discussed in the literature, I identify four factors that altogether form ‘the crisis’ and are suggested to negatively affect Greek parties’ pledge-fulfillment. These are summarized as the economic crisis and recession, the austerity policies, the room to maneuver, and the political crisis. Economic crisis and recession The economic crisis and the economic recession are expected to affect Greek pledge fulfillment in two ways: first, as an exogenous shock, the crisis might have forced Greek parties to abandon their previous campaign priorities (Stokes, 2001). Secondly, the economic recession and the bad economic conditions are generally considered among the main factors that explain differences in policy outputs (Huber & Stephens, 2001), and evidence suggests that they indeed matter for pledge fulfilment even in relatively good economic times (Thomson et al., 2017). This is so because a bad shape of the economy does not provide governments with the necessary economic resources which are necessary overall for implementing policies, but particularly relevant to socioeconomic domains (Thomson et al., 2017). The deep and prolonged period of economic recession in Greece is therefore considered to negatively affect pledge fulfillment. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 13 Austerity Austerity refers to policies aiming at reducing government budget deficits through public expenditure cuts and tax increases (Armingeon & Baccaro, p. 161-2). In the Greek context, austerity is expected to affect pledge fulfillment in two ways: first, because it worsened the economic conditions, further reducing therefore governments’ revenues, and secondly, because austerity policies target by default social welfare policies and public expenditure spending, and are therefore directly relevant to socioeconomic pledges, but also to other areas like labor market policies. Room to maneuver The presence of the international ‘Troika’ is considered to affect fulfillment because it limited Greek parties’ room to maneuver. Literature generally suggests that the more institutional constraints there exist, the less the likelihood for governments to implement desired policies (see Lijphart, 1999; Powell & Whitten, 1993; Thomson et al., 2017). An extra layer of institutional and financial constraints, like the ‘Troika’, is expected to hinder the ability of Greek parties to fulfill promises. Political crisis The political implications of the crisis on the Greek political system as discussed more above in broad terms pertain to a crisis of governance that includes frequent changes and exit of coalition partners, and early disruptions in governments’ terms. These might limit pledge fulfillment both because of the instability of the government formations, and because of the reduced period of governing, as research suggests that time resources matter and negatively affect pledge fulfillment rates (Thomson et al., 2017). Taken together, these pressures are considered likely to negatively affect pledge fulfillment in Greece during the studied period. In the research design below, I describe the elections/governments under investigation, present my hypotheses, and the method used to evaluate them. 3. Research Design In this chapter I describe the case, launch my hypotheses, present my methodology, and end with a section on the selection of the promises, the sources for the evaluation of the fulfillment, and an outline of the analyses. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 14 3.1. Case: elections/governments The conclusions lifted forward in Thomson’s et al. (2017) study are tested in two consecutive Greek elections/governments: the 2009 elections and the subsequent PASOK 2009-2012 government, and the June 2012 elections, and the ND coalition 2012-2015 government. The elections and the governments are chosen as ‘crisis’ cases, and are therefore representative in that they enable the evaluation of the program-to-policy linkage in a crisis setting, as well as the comparison between ‘stable’ democracies. The 2009 elections coincide with the signaling of the economic crisis and have been called ‘The Elections of the Great Recession’ (Nezi & Katsanidou, 2014), while the 2012 elections revealed the political consequences of the crisis, and have been characterized as ‘twin earthquake elections’ (Voulgaris & Nicolacopoulos, 2014). In addition, they are the latest concluded governments and provide an understanding of the most recent developments. The PASOK government is divided in the period from 2009 to 2011 where it rules in a one-party majority government, and in the 2011 to 2012 period, where a ‘transitional’ government of national unity including PASOK, ND, and LAOS8 under the technocrat Papademos takes over (Nezi & Katsanidou, 2014). This ‘transitional’ government had as a goal to implement decisions taken in a European level (see EU summit in Table 1), to vote for the second MoU, and to proceed to new elections (Vasilopoulou & Halikiopoulou, 2013). The ND- led coalition is split in the 2012-2013 period of a tripartite coalition between ND, PASOK, and DIMAR9, and the phase from 2013 to 2015, with only ND and PASOK. The formation of these coalitions was unique for Greek standards, not only because of single-party governments being the norm, but also because of the composition of the cabinets. PASOK and ND had long been rival parties, and DIMAR as a reform leftist party was ideologically diverse at least from ND (Vasilopoulou & Halikiopoulou, 2013). Yet, these parties converged in economic terms under the pro EU/MoU divide, which subsequently united them in political terms as well (Kriesi, 2015). Table 1 below summarizes the main political developments. Table 1. Main political events in Greece: 2009-2015 Date Events 4 October 2009 National elections; PASOK forms a single-party majority government. 23 April 2010 Papandreou announces the request of the first rescue loan from the EU and the IMF. 8 LAOS (Popular Orthodox Rally). A populist radical-right party (Mudde, 2007; Tsirbas, 2015) 9 DIMAR (Democratic Left). A SYRIZA-splinter party founded in 2010 (Tsirbas, 2015) Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 15 2 May 2010 Signing of the first MoU, which includes €110 billion bailout fund and.the first austerity package. 26-27 October 2011 EU summit. Greece's European lenders decide on the Private Sector Involvement (PSI), i.e. a 50 per cent bond 'haircut', on the condition of new austerity measures. 31 October 2011 Papandreou announces a plan to hold a referendum over the new demands. 9 November 2011 Papandreou steps down amidst domestic and international discontent over his referendum plan. 11 November 2011 Formation of a government of ‘national unity’ between PASOK, ND and LAOS, under the technocrat Lucas Papademos. February 2012 Signing of the second MoU without the participation of LAOS. Approval of a €130 billion bailout fund and new austerity measures. 11 April 2012 New elections are decided for May 6. 6 May 2012 National elections; no government formation. 17 June 2012 Second election round; formation of a grand coalition between ND, PASOK, and DIMAR. 23 June 2013 DIMAR disagrees over the closure of the National TV Broadcaster and leaves the government. New cabinet by ND and PASOK. 24 December 2014 Parliament fails to elect a new President. New elections for January 25, 2015. Note: Information taken by Vasilopoulou and Halikiopoulou (2013) and Tsatsanis and Teperoglou (2016). 3.2. Hypotheses This section presents my hypotheses, and they aim at testing the conclusions of Thomson et al. (2017) in the Greek context of crisis. Which conclusions by Thomson et al. (2017) are chosen to be assessed depends on the variation my cases/governments offer. However, I also include one hypothesis that applies only to the ND government and to the Greek case. Taking the four factors identified in the theoretical approach, my main initial focus is their overall impact on Greek pledge fulfillment. I begin by assessing the applicability of the overall conclusion of Thomson et al. (2017) that parties in ‘stable’ democracies fulfill a relatively and sometimes substantially high percentage of their promises. Following the discussion on the theoretical section, I suggest that the economic crisis, the austerity policies, the ‘troika’, and the political crisis, contribute to lower pledge fulfillment by Greek parties during the studied period compared to results raised by previous research both in international (Thomson et al., 2017), and in Greek (Kalogeropoulou, 1989) context. My first hypothesis is: Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 16 H1: Greek pledge fulfillment during the studied period of pressure is lower compared to other countries under ‘stable’ times. With the same analysis carried out for H1, I also test the next conclusion of Thomson et al. (2017). This relates to whether their main finding on power-sharing arrangements applies in a democracy under pressure. Thomson et al. (2017) find that single-party governments are significantly more likely to fulfill their promises than governments in coalitions. Sharing power is considered to affect fulfillment rates -in broad terms- because single-party governments have more control over policymaking (Powell & Whitten, 1993; for a broader discussion see Thomson et al., 2017). To what extent can this conclusion be applied on a crisis setting? Testing the above argument in the Greek crisis context is possible because there is variation in government type. However, while ND is a clear coalition, PASOK is divided in the first two- year period where it is a single-party government, and in the last phase where it participates in a grand coalition. For the purposes of this project, it is considered as a single-party government throughout its term, although this decision can also be considered problematic. Reasons for this are first, that for most of the period PASOK is governing alone; second, that the last period is overall a ‘depoliticized’ governing term in that the only scope was the implementation of decisions taken at a European level; third, that studies find that PASOK was considered the main coalition pillar (Nezi & Katsanidou, 2014); last, because it is not possible in the analysis to disentangle the two periods, and if a decision needs to be taken, PASOK is ‘closer’ to a single-party than a coalition government. In the context of crisis, two possible expectations can be addressed. On the one hand, one could expect the single-party government to fulfill more promises than the coalition, in line with the findings of Thomson et al. (2017). This is mainly because a coalition faces more constraints compared to single-party government, because it is forced to share power. The power-sharing arrangements are part of a broader discussion on institutional and other types of constraints and their impact on pledge fulfillment (Lijphart, 1999; Powell & Whitten, 1993; Thomson et al., 2017). In the Greek case, while both governments are under the same constraints/pressures of the ‘crisis’, a coalition ‘crisis’ government is subjected to more constraints that stem from being compelled to share power. Thus, one expectation could be that just like under times of stability, ‘crisis’ coalition governments fulfill fewer pledges than single- party ones. On the other hand, a second hypothesis could be that a coalition government under a crisis fulfills its promises to a satisfying extent. This assumption relates to the theoretical Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 17 discussion regarding the across-the-board political consensus among parties in times of crises (e.g. Lijphart, 1996; Bueno de Mesquita, 1981; Chowanietz, 2011). An implication of this for pledge fulfillment could be that a coalition ‘crisis’ government is able to fulfill its promises because of the ‘unity’ and the agreement that is established among parties. In that case, the constraints of sharing power are ‘cancelled out’ or balanced by the consensus agreement between parties. In the Greek case, the coalition between ND, PASOK, and DIMAR, although formed after the elections, was a grand coalition of ‘national salvation’ (Vasilopoulou & Halikiopoulou, 2013), that shared common goals and objectives. Moreover, another argument in line with this scenario relates to the institutional context within which pledges are made. It could be argued that parties potentially adjust their promises to the situation they expect to face when elected; that could be translated into parties making fewer, less difficult, or modest promises, something that also relates to the ‘cancelling out’ effect (Thomson et al., 2017). ND therefore could be expected to have made few and ‘realistic’ promises as it was elected amidst the crisis. The above considerations lead to the formulation of two hypotheses that aim to capture both the strong research evidence, and the theoretical discussion on coalitions under crises: H2a: The coalition ND government fulfills fewer pledges than the single-party PASOK government. H2b: There is no substantial difference on pledge fulfillment between the two governments. Furthermore, the next conclusion by Thomson et al. (2017) I examine relates to the characteristics of pledges in connection to their likelihood of being fulfilled. I focus on the promises that intend to preserve the status quo (see method), and those that intend to change it. Status quo pledges have been systematically found to have significantly higher rates of fulfillment compared to ‘change’ pledges, because of the incremental nature of policymaking (Thomson et al., 2017). If this finding ‘holds’ under crises as well, one hypothesis would be to expect similar findings in the Greek context as well. However, the fulfillment of status quo promises in a crisis context is not necessarily easy; previous research suggested that under crises, governments may be forced to implement large-scale reforms that contradict their intentions and go beyond incremental change (e.g. Kriesi, 2015). In addition, the common denominator of the theoretical discussion in the literature on crises and their effects, is that crises induce overall changes in the political or party system they occur (e.g. Walby, 2015; Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 18 Kriesi, 2015). In the Greek context, the imposition of austerity reforms by the ‘Troika’ on the one hand, and the broader changes that occurred in terms of party system change or policy innovations might indicate that preserving the status quo is not easier than changing it. Taking these recommendations into account, I formulate two further hypotheses that aim to include these contradictory expectations: H3a: Pledges that promise to preserve the status quo are more likely to be fulfilled than pledges that promise to change it. H3b: Pledges that promise to preserve the status quo less likely to be fulfilled than pledges that promise to change it. In addition, I further investigate ND’s promises separately, to examine their fulfillment in relation to the need for negotiation with the troika. ND in its program included ’18 main points’, out of which 10 would be pursued after negotiating with the lenders, while the remaining 8 without negotiation. Following again the discussion on constraints, I suggest that the promises that need negotiation will be less fulfilled than those that do not. My fourth hypothesis is: H4: ND’s pledges that require negotiation with the ‘Troika’ are less likely to be fulfilled than those that do not. Last, I examine the type of ‘change’ pledges in relation to their fulfillment. Especially in the context of the crisis, pledges that involve more resources from parties for their fulfillment are arguably less likely to be fulfilled. The analysis begins by exploring the making of Greek pledges, including both the total amount of pledges made by PASOK and ND, and the promises evaluated for fulfillment. Next, I continue by assessing the hypotheses 1, 2a and 2b by looking at the total fulfillment results for both governments. Then I examine the hypotheses 3a and 3b by looking separately the fulfillment of change/status quo pledges; with the same analysis I also examine the fulfillment of the type of ‘change’ pledges. Finally, I end with the separate examination of ND’s promises for H4. In the method section below, I describe the methodology applied to answer these hypotheses. 3.3. Method Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 19 This section presents the method used for the study of Greek election pledges. I start by describing the general type of method used, and then I continue with the pledge approach, starting with the definition and categorization of pledges, and continuing with the criteria for the evaluation of their fulfillment. Then I describe the limitations of the method, and at the end I briefly outline the analyses. The general methodology I use for this study is quantitative content analysis (Riffe, Lacy & Fico, 2005). 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, 2013, p. 24). One of the most crucial and necessary criteria that need to be at place when conducting a content analysis is the reliability and replicability of the results. This means that by applying the same procedure, different scholars, at different times and situations, should be able to produce the same outcomes (Krippendorff, 2013, p. 24). Quantitative content analysis refers to ‘the systematic and replicable examination of symbols of communication, which have been assigned numeric values according to valid measurement rules and the analysis of relationships involving those values using statistical methods, to describe the communication, draw inferences about its meaning, or infer from the communication to its context, both of production and consumption’ (Riffe, Lacy & Fico, 2005, p. 25). In my case for example, this means that a text is coded and transformed from a qualitative to a quantitative form by using specific coding principles and statistical software. Below I describe more specifically the process for the content analysis and coding of election manifestos. The specific methodology I use in this project for the analysis of Greek pledges follows the overall principles and method by Thomson et al. (2017). The 12 scholars coordinated their previously independently undertaken studies in a comparative project with common conceptualizations, and reliability and robustness tests, that produced cross-nationally comparable data and framework (Thomson et al., 2017). Detailed definitions and previous scholarly approaches are found in the online supporting information that accompanies their publication. I chose to use this method because my aim is to compare my findings to Thomson’s et al. (2017), therefore is the most suitable and preferable choice. In addition to the overall coding principles mentioned above, the specific categorization schemes are taken from the technical report of the project ‘If Elected, I Promise…’ by Elin Naurin, Nicklas Håkansson, and Björn Werner (2014; report’s version: April 2015), which is available upon request. Moreover, I have further added and changed some of the coding to make them more suitable for my data and cases, which I specify in the relevant section. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 20 The research approach in this project includes three main steps: first, a content- analysis of election manifestos to identify and categorize pledges, second, the evaluation of their fulfillment by qualitatively assessing policy actions and outcomes (Thomson et al., 2017), and third a qualitative comparative analysis with the findings of Thomson et al. (2017). 3.3.1. Definition of pledges The initial criterion for a statement to count as a pledge is that it includes commitment language for a future action or outcome that indicates unequivocal support on behalf of the party. As long as the commitment is clear, the firmness of the formulation is not restrictive, therefore both ‘soft’ (‘we want to’) and hard (‘we promise’, ‘we will’) pledges are included. The final decision over which statement qualifies as a promise is based on the so-called testability criterion: a commitment to an action or outcome needs to be able to be tested through evidence that confirms or rejects the eventual fulfillment (Thomson et al., 2017). This means that a clear strategy should be provided either 1) by the manifesto writers, by clearly offering the way toward testing fulfilment, or 2) by the scholars themselves, through a logic that suggests what is relevant to examine. These two criteria lead to a distinction between specific (criterion 1) and vague (criterion 2) pledges (Thomson et al., 2017). I use both in my coding, but I only examine the specific pledges for fulfillment. In this thesis I follow Thomson et al.’s (2017, p. 6) definition of an election promise as ‘a statement committing a party to one specific action or outcome that can be clearly determined to have occurred or not’. Using a common definition facilitates the comparability of my findings with their conclusions. When categorizing promises, the unit of analysis is the pledge; this means that promises can differ in their length and structure, may consist of more than one sentence, and/or appear with different wordings and at several times in the manifesto. Regarding the latter, when an exact same pledge is repeated multiple times, it is counted as one pledge, while when it is stated multiple times but in different degrees of precision, the most specific formulation counts (Thomson et al., 2017). Pledges in ‘packages’, i.e. statements that include multiple propositions together, are sometimes treated as one pledge, and sometimes are split in separate pledges (Naurin, Håkansson & Werner, 2014). Commitments in ‘packages’ that are implied to be fulfilled by a single political measure, are not considered individual promises. For example, ND’s promise to ‘Restore the special pay scale scheme for police officers that remain in service and of the Air Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 21 Force pilots’ falls into this category. In contrast, when pledges are presented together, but as part of a larger aim or common context, they need to be tested and categorized differently. For example, ‘Our goal is the further development of the subway, the tram, the railway and the suburban railway…’ (PASOK, 2009). The cases where a statement does not constitute a pledge refer to sentences that fail the testability criterion. These include either ‘descriptions of reality’, or ‘rhetorical statements’ about the future. The former refers to discussions of the party’s accomplishments, the country’s situation, or other parties’ actions. Rhetorical statements refer to statements that are either too vague, or aim at creating emotions, and therefore are not testable (Naurin, Håkansson, & Werner, 2014). Last, statements that although testable, promise something that is about to take place after the end of the government term, are excluded from the coding. An example by PASOK is: ‘We aim to reduce greenhouse gas emissions by 65% by 2050’. 3.3.2. Coding of pledges Pledges are coded according to their content in the sub-categories of issue area, type of change, action and outcome, and for ND in an extra category of ‘negotiation/without with the troika’. In this part I am inspired by the technical report from Naurin, Håkansson, and Werner’s study on Sweden (2014), but have coded some pledges differently, which I specify. The overall coding of the pledges is found in Appendix 1 and 2. First, the pledges are categorized into policy areas. The coding scheme by Naurin, Håkansson, and Werner (2014) which I follow includes the categories of agriculture, economics, enterprise, employment-labor market, social welfare, migration, education, infrastructure, legal matters, European Union, environment, culture, foreign policy. To these, I added the category ‘political system/public administration’ (see Appendix 2 for details). When two or more issue areas seem suitable, the one associated with the ultimate imperative is chosen. This means that PASOK’s promise to ‘Increase the VAT refund rate for farmers from 7% to 11%’ is coded under ‘agriculture’, not ‘economics’. This decision is primarily taken with regard to the consistency and reliability of the coding. Secondly, pledges are categorized depending on the type of change that is promised, and particularly on their intention to preserve or change the status quo. The change category is divided into subcategories of change which include increases or cuts in spending, increases and cuts in taxes, and ‘other’ changes which is a broad category where all other types of change that do not fit in the previous ones fall into. There is also a last category of review Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 22 pledges, which refers to promises for an investigation of some issue. Details and examples are found in Appendix 2 (Naurin, Håkansson, & Werner, 2014). Third, pledges are distinguished into action and outcome pledges. Action promises are statements where the party specifies or implies the way for reaching a goal. An outcome pledge promises a specific measurable result, but the actions for achieving it cannot be disentangled. For instance, a statement to ‘increase employment’ is an outcome promise, but if the party detailed how to increase it, it would be an action pledge. The action or outcome does not relate to specificity or vagueness, and action and outcome pledges can be both specific and vague (Thomson et al., 2017). For example, ND’s promise ‘Extension of the unemployment benefit from 1 to 2 years’ is a specific action pledge, while ‘We promise to create 150.000 new jobs in the private sector by mid-2013’ is a specific outcome promise, because the goal is clear, but the means to achieve it are not clarified. Last, ND’s pledges are further categorized according to whether they would be pursued after negotiation deliberations with the ‘Troika’ or not. In its program, the party leader separated its promises according to this distinction, therefore the criteria for the classification are offered by the party. 3.3.3. Evaluation of fulfillment In line with previous pledge research, I code the fulfilment of pledges using three categories: fulfilled, partially fulfilled and not fulfilled (Thomson et al., 2017). Fulfillment is determined differently for different types of pledges. Action promises are fulfilled if relevant action is taken in line with what was promised, while outcome promises are fulfilled it the outcome is produced. In addition, the action or outcome must be taken or produced within the end of the election term for a pledge to be fulfilled. An exception to this relates to if the party sets a specific threshold itself, which I discuss below (Thomson et al., 2017). Status quo pledges are fulfilled when the policy area remains unchanged. Pledges are coded as partially fulfilled first, if there is some action taken, or some outcome produced, but not equivalent to what was promised. Secondly, if the party sets a specific period until when the promise will be fulfilled, but the fulfillment takes place after this period, but still within its government term, then it is coded as partially fulfilled. This category is important because it accounts for the modern political environment that does not always allow parties to achieve their desired goals in the expected way (Naurin, Håkansson, & Werner, 2014). Last, pledges are considered broken in three occasions: first, if no significant action is made, regardless of whether it was the party’s fault, or due to the existence of various Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 23 constraints. Secondly, if there have been efforts, intentions, or symbolic actions toward fulfillment but without any sufficient result, or if legislation is introduced but failed to be realized. Third, specific pledges for which no information is found are also coded as broken (Thomson et al., 2017). 3.3.4. Comparative analysis and limitations The findings of my project are qualitatively compared to those by Thomson et al. (2017). I do that by resulting reports of their results on the equivalent analysis I conduct, which has also been done in other studies (see Costello & Thomson, 2008). Alternatively, and ideally, the comparative analysis should be conducted quantitatively, by collapsing my dataset to that of Thomson et al. (2017) in order to gain better insight and more intuitive results, and also to be able to explain variation in pledge fulfillment and control for other factors that could alter or condition my findings. This is an important limitation of this study, and is further discussed below. Yet, since this project aims to offer a first insight on pledge fulfillment under crises and in Greece, the qualitative comparison facilitates this initial exploratory scope. An important limitation of my project concerns the reliability of the coding and the results. As mentioned above, reliability and replicability of the findings is crucial in all scientific research. Thomson et al. (2017) conducted inter-coder reliability tests for their comparative project, and the country-focused studies also performed such tests using native- speakers to code parallelly the manifestos. In this project such tests were not taken; ideally, more than one Greek native speakers should have conducted the analysis with the same principles and coding schemes as me, and the agreement results between us should have exceeded certain thresholds of reliability scores in order for this study to be reliable (Riffe, Lacy & Fico, 2005; Krippendorff, 2013). Such an option was not possible however. While I studied thoroughly the methodology guidelines and the overall principles, and did a systematic work repeatedly to eliminate mistakes, the limitations are not overcome, and the reliability is not enhanced; further analyses should be carried out at a later stage to ensure reliability and replicability. 3.4. Selection of promises and sources for fulfillment Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 24 In this study I code all pledges made by PASOK and ND prior to the 2009 and June 201210 elections accordingly, which results in a unique dataset of 652 pledges. For the 2012-2015 ND coalition, I only include the promises made by ND, and not its coalition partners PASOK and DIMAR. This is decided both because ND was the main party in the government, with the largest share of the votes and the majority of the members in ministerial positions11, but also because of the limitations of this project that did not allow for the coding of two extra manifestos. PASOK’s promises are derived from its official election manifesto titled ‘The citizen first’ (PASOK, 2009). ND did not issue an official program, but instead the party leader announced his programmatic promises in the media. Both documents are retrieved from the Manifesto Corpus database (Lehmann et al., 2017), but for ND’s pledges I further consulted newspaper articles and listened to the party’s leader speech to confirm them, which resulted in the addition of five more promises to the ones found in the Manifesto Corpus (Lehmann et al., 2017). The dataset, as well as a file including the documentation of the fulfillment are available upon request. While all pledges are coded, those evaluated for fulfillment include all of ND’s specific pledges, and PASOK’s most important specific pledges. The decision over which of PASOK’s promises are the most important is facilitated by the party’s manifesto. This decision is based first, on the fact that these pledges dominated the 2009 elections and received most media coverage, and secondly because of the large number of PASOK’s pledges, that make the analysis difficult for the limitation of this study. However, it could be argued that these promises are ‘biased’ in some ways; on the one hand, because they are the party’s most important promises, and since they were emphasized, it could be reasonable that it fulfilled them. On the other hand, because these promises focus primarily on the economy, it could be possible that they were not redeemed because they were directly linked to the crisis. Yet, this decision is justified for two reasons: first, although these promises refer overall to the economy, they do not entail only increases in spending, but other changes as well, that are not directly relevant to resources. In addition, many other domains and promises require expenses and spending, therefore the ‘bias’ might apply to other categories as well. Secondly, analyzing a party’s most important pledges facilitates a limitation of pledge research: the fact that not all 10Prior to the May 2012 elections, ND gave two programmatic speeches instead of a manifesto, but prior to the June 2012 elections made only these ‘18 points’. I only analyze June’s promises since these elections led to the formation of the 2012-2015 government. 11 The initial composition of the government included 39 members, among which 3 belonged to PASOK and 2 to DIMAR. Election results: ND: 29,66%, PASOK: 12,28%, DIMAR: 6,26%. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 25 promises have the same importance, which has as a result that when analyzing a party’s total amount of pledges, the overall conclusions do not shed light on the fate of promises that concerned the electorate most (Thomson et al., 2017). The fulfillment of Greek pledges is evaluated using the following sources: 1) The Hellenic Parliament website (2017) [Βουλή των Ελλήνων]) where all legislative work is available. 2) The governmental program ‘Diavgeia’ (2017) [Διαύγεια, ‘Transparency, Clarity’] where all decisions by government bodies and public administration agencies are published online. 3) The Government Gazette (2017) [Εφημερίδα της Κυβερνήσεως], which is the official journal of the Government of Greece that lists all laws, legislation, presidential decrees and ministerial decisions ratified by Cabinet and President, offering also a search engine with key words. For outcome promises, statistical data from the Hellenic Statistical Authority [Ελληνική Στατιστική Αρχή] are consulted. Last, in order to capture the background of legislation, but also to disentangle some hard cases, search engines like google are also used. 4. Results This chapter presents the findings, evaluates the hypotheses, and compares the results with those of Thomson et al. (2017). 4.1. Pledge-making in Greece The analysis begins with a discussion of the amount and types of pledges made by Greek parties during the studied period. I start with both parties’ total pledges, to offer a general understanding of Greek pledge making, and to show the analogy of PASOK’s most important promises compared to their total amount. Then I focus on the pledges analyzed for fulfillment. Table 2 below offers a summary of all pledges. Table 2. Total number of pledges made in PASOK’s and ND’s election programs per category. Categories PASOK (2009) ND (June 2012) BOTH Specific 491 (81%) 43 (93,5%) 534 (81,9%) Vague 115 (19%) 3 (6,5) 118 (18,1%) Action 598 (98,7%) 43 (93,5%) 641 (98,3%) Outcome 8 (1,3%) 3 (6,5%) 11 (1,7%) Status quo 42 (6,9%) 6 (13%) 48 (7,4%) Change: cut 3 (0,5%) 0 (0%) 3 (0,5%) Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 26 Change: expense 116 (19,1%) 15 (32,6%) 131 (20,1%) Change: tax cuts 7 (1,2%) 12 (26,1%) 19 (2,9%) Change: tax increase 6 (1%) 0 (0%) 6 (0,9%) Change: other 412 (68%) 13 (28,3%) 425 (65,2%) Review 20 (3,3%) 0 (0%) 20 (3,1%) Agriculture 22 (3,6%) 1 (2,2%) 23 (3,5%) Economics 45 (7,4%) 24 (52,2%) 69 (10,6%) Enterprise 31 (5,1%) 5 (10,9%) 36 (5,5%) Employment, labor market 26 (4,3%) 8 (17,4%) 34 (5,2%) Social welfare 80 (13,2%) 3 (6,5%) 83 (12,7%) Migration 37 (6,1%) 2 (4,3%) 39 (6%) Education & research 46 (7,6%) 0 (0%) 46 (7,1%) Infrastructure 53 (8,7%) 0 (0%) 53 (8,1%) Legal matters 32 (5,3%) 2 (4,3%) 34 (5,2%) EU 13 (2,1%) 0 (0%) 13 (2%) Environment 66 (10,9%) 0 (0%) 66 (10,1%) Culture 65 (10,7%) 0 (0%) 65 (10%) Foreign policy 50 (8,3%) 1 (2,2%) 51 (7,8%) Political system / Public administration 40 (6,6%) 0 (0%) 40 (6,1%) Total (N) 606 (100%) 46 (100%) 652 (100%) Note: Total number of pledges made by PASOK and ND per categories. Percentages in parentheses. The right column shows total number and percentages for both parties. The most striking finding from table 2 relates both to the large difference in the amount of pledges made by PASOK and ND, as well as to the difference in the emphasized areas. PASOK makes a total of 606 promises, dispersed across all relevant categories, indicating its ability to offer voters a detailed and specific governing program, and to make promises not only in central areas but also in peripheral ones. Its most emphasized issue areas are the ‘social welfare’ category (13,2%), followed by ‘environment’ (10,9%), and ‘culture’ (10,7%). ND on the other hand makes a significantly less amount of promises, 46 in total, and the policy domains are limited. The majority of ND’s pledges are in the economy (52%), with employment/labor market (17,4%), and enterprise (10,9%) following. The rest of ND’s promises refer mainly to legal issues and migration (4,3% on both), revealing a very economy-centered election program, with a secondary focus on ‘security’. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 27 Regarding the ‘change’ pledges, both parties’ majority of promises falls into the ‘other’ category (65,2%), which includes a diverse range of promises, from changes in public administration, to changes in the business environment or legal matters. The second most common category is the ‘expense’ (20,1%), which includes higher public spending for a variety of domains, increases in salaries, or larger unemployment benefits. PASOK’s promises that require public spending for their fulfillment is 19,1 percent, whereas ND’s 32,6 percent. This is a very interesting finding, because previous studies show that ‘expense’ pledges during times of economic crises or recession tend to be quite fewer than during times of growth (Thomson et al., 2017), and especially for ND we see that the percentage of ‘increase’ pledges is quite high. Another interesting finding refers to the pledges that promise to preserve the status quo, which include promises such as ‘no further reduction in salaries’, or ‘no new taxes’. In a 7,4 percent of their pledges, both parties promise to maintain the status quo, which is the third largest category after ‘other’ and ‘expense’. PASOK’s status quo promises are 6,9 percent, while there is a big difference with ND (13%). These percentages and especially ND’s are considered high compared to other cases (Thomson et al., 2017). Last, both parties’ majority of pledges are specific (81,9%), where there is only one way to fulfillment, showing that they are able to make specific promises, in line with overall findings (Thomson et al., 2017). Action promises are also the majority among Greek parties’ pledges, being 98,3 percent for both. The study of Greek parties’ pledges shows that PASOK has a much more analytic election program that covers all issue areas, but ND presents more limited options to voters, although it includes overall salient areas like socioeconomic pledges. Moving on, Table 3 below provides a summary of the pledges examined for fulfillment. The discussion below mainly involves PASOK’s promises, since ND’s are almost the same as above, with the exclusion of only 3 vague promises. Also, it should be noted that the initial number of PASOK’s most important specific promises was 83, but six of them were left out because of contradictory information on their fulfillment. More specifically, while evidence from many journalistic articles suggested that they were fulfilled, it was not possible for me to find official evidence, potentially due to time limitations. In order to avoid a wrong judgment, I left six of them aside without categorizing them as not fulfilled. Table 3. Number of pledges analyzed for fulfillment per party and category. Categories of pledges PASOK (2009) ND (June 2012) BOTH Action 74 (96,1%) 41 (95,3%) 115 (95,8%) Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 28 Outcome 3 (3,9%) 2 (4,7%) 5 (4,2%) Status quo 4 (5,2%) 6 (14%) 10 (8,3%) Change: cut 2 (2,6%) 0 (0%) 2 (1,7%) Change: expense 10 (13%) 15 (34,9%) 25 (20,8%) Change: tax cuts 4 (5,2%) 12 (27,9%) 16 (13,3%) Change: tax increase 6 (7,8%) 0 (0%) 6 (5%) Change: other 51 (66,2%) 10 (23,3%) 61 (50,8%) Agriculture 2 (2,6%) 1 (2,3%) 3 (2,5%) Economics 34 (44,2%) 24 (55,8%) 58 (48,3%) Enterprise 20 (26%) 5 (11,6%) 25 (20,8%) Employment, labor market 5 (6,5%) 8 (18,6%) 13 (10,8%) Social welfare 2 (2,6%) 3 (7%) 5 (4,2%) Migration 0 (0%) 1 (2,3%) 1 (0,8%) Infrastructure 3 (3,9%) 0 (0%) 3 (2,5%) EU 1 (1,3%) 0 (0%) 1 (0,8%) Environment 1 (1,3%) 0 (0%) 1 (0,8%) Foreign policy 0 (0%) 1 (2,3%) 1 (0,8%) Political system / Public administration 9 (11,7%) 0 (0%) 9 (7,5%) Total (N) 77 (100%) 43 (100%) 120 (100%) Note: All pledges tested for fulfillment per category and party. Table shows PASOK’s most important specific pledges, and ND’s specific pledges. Right column presents the total numbers and percentages for both parties. The pledges presented above include all the specific pledges for ND, and the specific most important pledges for PASOK. In its 2009 manifesto, PASOK prioritized a ‘100-days crisis recovery plan that will tackle directly all the big problems and will set the bases for the big changes’ (p. 10). The promises included here are those included in this part of PASOK’s program. The evidence in table 3 shows a clear focus on the economy: 44,2 percent of the promises are made within ‘economics’, followed by 26 percent in ‘enterprise’ and 11,7 percent in ‘political system/public administration’. Compared to the percentage of its economic pledges out of the total amount of pledges (7,45%), it is evident that the majority of them are included in its most prioritized pledges. Regarding the change categories, most of PASOK’s most important promises are in the ‘other’ coding (66,2%), followed by ‘expense’ (13%) and tax increase (7,8%). Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 29 To give a complete picture of both parties’ promises analyzed for fulfillment, these fall mainly into ‘economics’ (48,3%), ‘enterprise’ (20,8%), ‘employment-labor market’ (10,8%), ‘public administration’ (7,5%) and ‘social welfare’ (4,2%). The majority of change promises are ‘other’ (50,8%), followed by ‘expense’ (20,8%), and ‘tax cuts’ (13,3%), while the pledges to preserve the status quo are 8,3 percent in total. 4.2. Pledge fulfillment in Greece In this section I present the findings of the fulfillment of Greek pledges, evaluate my hypotheses, and compare the results with Thomson’s et al. (2017). I begin with a discussion of the results for both parties and the evaluation of H1, and H2a and H2b, and I compare them with Thomson’s et al. (2017) findings. At a second stage, I evaluate H3a and H3b by presenting the results for the change/status quo pledges, and last, I present ND’s pledge fulfillment separately to examine H4. Table 4 below summarizes the main findings on the fulfillment of PASOK’s and ND’s pledges. Table 4. Fulfillment of pledges in Greece. (Status) PASOK (2009) ND (June 2012) BOTH Fulfilled 8 (10,4%) 7 (16,3%) 15 (12,5%) Partially fulfilled 59 (76,6%) 10 (23,3%) 69 (57,5%) At-least partially fulfilled 67 (87%) 17 (39,6%) 84 (70%) Not fulfilled 10 (13%) 26 (60,5%) 36 (30%) Total N 77 (100%) 43 (100%) 120 (100%) Note: Fulfillment results for PASOK and ND. The ‘at-least-partially fulfilled’ category includes the cumulative percentages of ‘fulfilled’ and ‘partially fulfilled’, and is highlighted with bold letters. The results in table 4 show the amount of pledges that are fully, partially, and not fulfilled by party and in total. Taken together, the two parties fully fulfilled 12,5 percent of their promises, partially fulfilled 57,5 percent, and did not fulfill 30 percent. If the full and partial fulfillment are collapsed, the two governments score 70 percent of at-least partially fulfilled promises. The most striking finding is the large difference between the two parties: PASOK fulfills at-least partially 87 percent of its promises, while ND ‘breaks’ the majority of its promises by 60,5 percent. The findings do not provide clear support for my first hypothesis that suggested that Greek pledge fulfillment during the studied period will be lower compared to Thomson et al. (2017), and to Kalogeropoulou (1989). PASOK during the 1981-1985 government fulfilled Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 30 70,5 percent of its manifesto promises (Kalogeropoulou, 1989, p. 297), and the governing parties in Thomson et al. (2017) fulfilled on average between 60 and 90 percent of their promises at-least partially (p. 8). My findings show that both parties fulfilled a total of 70 percent of their promises at-least partially, which is a quite high percentage, and compares very well with other countries, being also much higher than the lowest 60 percent in Thomson et al. (2017). The overall results are at odds with the expectations, therefore if both parties’ percentages are taken into account as a whole, H1 is not supported. However, the large differences between the two parties necessitate a separate comparison; regarding PASOK, its 87 percent of at-least partial fulfillment is even higher than its previous results, and close to the highest rates observed in ‘stable’ democracies (Thomson et al., 2017). By looking at PASOK only, H1 is also rejected. The results of ND on the other hand support the hypothesis, as ND’s 39,6 percent of at-least partial fulfillment is 20 percent lower from the lowest 60 of the governing parties in Thomson et al. (2017) and by 30 percent compared to Kalogeropoulou (1989). H1 can be considered confirmed only for ND, and is rejected for PASOK as well as for the whole period under examination. My second hypotheses set out to test whether Thomson’s et al. (2017) finding about power-sharing ‘holds’ in a democracy under pressure. They find that single-party governments fulfill a significantly higher amount of pledges than governments in coalitions (p.10-11). Since the expectations for governments under crises were conflicting, I claimed that there are both reasons in favor (H2a) and against (H2b) this assumption in a crisis context. The findings from the above analysis clearly support the conclusion by Thomson et al. (2017), since the ND coalition at-least partially fulfilled a significantly lower amount of promises than the single-party PASOK government. Therefore, H2a is confirmed and H2b rejected. An example of PASOK’s fully fulfilled promise is the provision of an exceptional and temporary solidarity allowance for the most vulnerable groups (law 3808/2009). PASOK’s partially fulfilled promises are mainly pledges implemented after the three-month threshold set by the party. Such an example is the application of a progressive, indexed tax scale on all incomes (law 3842/2010). Its broken pledges are for example promises to raise salaries and pensions above inflation, or increase the unemployment benefit to the 70% of the basic income. An example of ND’s fully fulfilled promise is the reduction of an exceptional and temporary levy introduced by PASOK in 2009 (law 4305/2014); a partially fulfilled pledge is the reduction of the VAT for the catering sector from 23 to 9 percent (reduced to 13 instead of 9, laws 4172/2013 & 4224/2013), while a broken pledge is the promise to reduce the high rate of VAT from 23% to 19%, the medium from 19% to 9%, and the low from 6,5% to 5%. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 31 The analysis continues with testing H3a and H3b regarding the fate of status quo pledges compared to the ‘change’ promises. Table 5 below presents the results for the two parties. Table 5. Pledge fulfillment by PASOK and ND per status quo/change. PASOK ND BOTH At-least partially fulfilled Not fulfilled At-least partially fulfilled Not fulfilled At-least Not fulfilled partially fulfilled Status quo 2 (50%) 2 (50%) 0 (0%) 6 (100%) 2 (20%) 8 (80%) Change_cut 2 (100%) 0 (0%) 0 (0%) 0 (0%) 2 (100%) 0 (0%) Change_expense 6 (60%) 4 (40%) 7 (46,7%) 8 (53,3%) 13 (52%) 12 (48%) Tax cuts 4 (100%) 0 (0%) 4 (33,3%) 8 (66,7%) 8 (50%) 8 (50%) Tax increase 6 (100%) 0 (0%) 0 (0%) 0 (0%) 6 (100%) 0 (0%) Other 47 (92,2%) 4 (7,8%) 6 (60%) 4 (40%) 53 (86,9%) 8(13,1%) All 67 (87%) 10 (13%) 17 (39,6%) 26(60,5%) 84 (70%) 36 (30%) Note: ‘At-least partially’ and ‘not fulfilled’ pledges per party and change/status quo. ‘At-least partial fulfillment’ with bold letters. To read this table: the second column called ‘at-least partially fulfilled’ indicates that PASOK fulfilled at least partially 60% of its ‘expense’ promises, or 50% of its ‘status quo’ promises. Column to the right shows total percentages of at-least partial and not fulfillment per each category for both parties. Regarding the status quo pledges, table 5 shows that PASOK and ND fulfilled them by only 20 percent at-least partially. PASOK fulfilled half of them, while ND did not fulfill any of them. Compared to all the other ‘change’ categories, the ‘status quo’ presents the lowest percentage. My hypotheses suggested that on the one hand, status quo pledges will be more likely to be fulfilled than ‘change’ pledges, following previous systematic evidence (H3a), while H3b countered that under the crisis and the various pressures, the pledges regarding the maintenance of the status quo would be less likely to be redeemed than those that aim to change it. Comparing the fulfillment of the ‘status quo’ with that of ‘change’ pledges, the results indicate that change of any kind is much more likely to occur than preserving the current state of affairs, as the ‘status quo’ pledges are the least fulfilled among all categories. Both the findings for the results of the status quo pledges separately, and their comparison to the ‘change’ pledges are impressive in connection to the conclusions from ‘stable’ democracies. Thomson Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 32 et al. (2017) find that of the 871 status quo pledges made by governing parties, 768, or 88%, were at-least partially fulfilled (p. 10), while these pledges are much more likely to be fulfilled in comparison to the promises of ‘change’. The findings therefore provide strong support for H3b, and H3a is rejected. The results further support the sub-expectation that among the ‘change’ pledges, those that require more recourses from the parties are less likely to be redeemed. The total results from both parties show that pledges to increase expenses and cut taxes were the least fulfilled, with 52 and 50 percent accordingly, which both require more resources. The highest percentage of at-least partial fulfillment by both parties are pledges aiming at cutting expenses (100%) and increasing taxes (100%), which are also in line with the character of the austerity policies, and the next category is ‘other’ types of pledges (86,9%) which is reasonable since they do not require spending. Finally, the last hypothesis relates to ND’s pledges and their fulfillment depending on the negotiation with the ‘Troika’. Five of the promises have been excluded because it was unclear how they would be pursued. Table 6 below presents the results. Table 6. Fulfillment of ND’s pledges: negotiation with the ‘Troika’. With negotiation Without negotiation Fulfilled 2 (7,4%) 5 (45,5%) Partially fulfilled 6 (22,2%) 4 (36,4%) At-least partially fulfilled 8 (29,6%) 9 (81,8%) Not fulfilled 19 (70,4%) 2 (18,2%) Total 27 (100%) 11 (100%) Note: ND’s pledge fulfillment according to negotiation with the ‘Troika’. At-least partial fulfillment with bold letters. Five (5) pledges are excluded because it was unclear in the party’s program how they would proceed. Total number pf pledges here: 38. Table 6 shows the fulfillment rates for ND’s promises categorized by the need for negotiation with the ‘Troika’. Hypothesis 4 expected that pledges that are pursued after negotiation are less likely to be fulfilled than those that are put forward without. The evidence clearly shows that pledges that required negotiation have significantly lower fulfillment rates; ND fulfilled 29,6 percent of the promises after negotiation at-least partially, while did not fulfill 70,4 percent. On the contrary, the party implemented 81,8 percent of its promises at-least partially without negotiation, and did not follow through on 18,2 percent. These findings provide further support Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 33 for the important role of institutional constraints on limiting incumbent parties’ pledge fulfillment. Judging from the above, my hypothesis is considered confirmed. To sum up my results, the main hypothesis (H1) regarding Greek pledge fulfillment being lower compared to other countries is partly supported, hypotheses H2a, H3b and H4 concerning accordingly the power-sharing arrangements, the ‘status quo’ compared to the ‘change’ pledges, and the ND’s promises in relation to the ‘Troika’ are considered confirmed, and hypotheses H2b and H3a are rejected. The discussion of the results, and the several issues that need to be addressed are analyzed in the following sections. 5. Discussion This section begins with a discussion of the main findings and continues with the limitations of this project and suggestions for alternative and further operationalizations. This thesis primarily aimed to determine the strength of the program-to-policy linkage in Greece, chosen as a democracy is crisis, and to assess the extent to which the suggestions derived from ‘stable’ democracies apply in such a context. The questions this project raised related to whether normative expectations of well-functioning democracies (e.g. Downs, 1957), as well as empirical findings that confirm these suggestions (Thomson et al., 2017) can be met by governing parties in a political system under the pressures of a deep economic crisis, by concentrating particularly on Greece. The suggested expectation was that under the various constraints, Greek parties would achieve lower rates of pledge fulfillment compared to findings in other countries, as well as in comparison to previous Greek results (Kalogeropoulou, 1989). The findings of this study provide mixed evidence. The overall 70 percent of at- least partial fulfillment by both governments during the studied period is considered as a relatively high rate under any circumstances. Especially PASOK’s 87 percent is against all odds in relation to what was hypothesized. On the contrary, ND’s results are moving toward the suggested direction. Below I elaborate on the results that found theoretical support, on those that were rejected, and on some issues regarding the variation between the two governments. ND’s 39,6 percent of at-least partial fulfillment is much lower than the lowest scores reported in Thomson et al. (2017), confirming that way the expectations lifted forward in the literature about representative democracies under crises, and my hypotheses. Although the general analysis has not been able to, nor focused on providing explanations on the exact factors that led to this result, it has been able to present more clear support for some. First, the Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 34 limited room to maneuver (Hellwig, 2001; Kosmidis, 2014), operationalized in the presence of the ‘Troika’ and examined in the fourth hypothesis, was found to matter significantly for pledge fulfillment variation among ND’s promises. In addition, the analysis has also confirmed expectations on power-sharing (Thomson et al., 2017; Powell & Whitten, 1993), and rejected suggestions on broad consensus among parties (e.g. Lijphart, 1996; Chowanietz, 2011). Although no other factors were controlled, the coalition ND’s results were in line with findings from ‘stable’ democracies that coalition governments fulfill fewer pledges than single-party cabinets. Moreover, the results have provided important evidence on the impediments to preserve the status quo in conditions of increased pressure. The results revealed that any form of change is much more likely to occur than any effort to maintain the state of affairs under crises. These findings bear great importance in the light of previous research (Thomson et al., 2017), as it is the first study to report such outcomes. Theoretically, this claim has been supported by accounts suggesting that crises are periods of overall changes (e.g. Walby, 2015), and of large-scale reforms that disrupt the incremental nature of ‘normal’ policymaking (e.g. Kriesi). Last, the theoretical discussion on political crises (Mainwaring et al., 2016; Kriesi, 2015), can be also considered as potentially confirmed if the judgement is made focusing on ND’s results, which was the government term mostly affected from how the crisis played out in electoral terms. Moreover, the analysis is considered to have rejected some of the theoretical propositions raised in the literature. PASOK’s results, although overall reject all the expectations, particularly seem to oppose the ones relating to the exogenous shock of the economic crisis causing ‘policy switches’ (Stokes, 2001). These suggestions mainly pertained to PASOK’s government, as it was the one elected prior to the official emergence of the crisis, therefore might have needed to change course. Finally, the overall percentage by both parties, although driven by PASOK’s results, can be considered as not providing strong overall evidence on the addressed expectations. However, the large difference among the two parties needs a closer look. While the lack of quantitative analyses did not allow to explain the large variation between PASOK and ND, one possible explanation might relate to the conditions of the crisis the two parties faced. The period under examination is not a ‘unified’ period in terms of the crisis. Rather, the crisis progressively deepened during time; economic conditions got worse, the austerity measures became harsher, and the political implications of the crisis deteriorated (e.g. Tsirbas, 2015). Therefore, although both governments were overall subjected to the same pressures, PASOK could be considered as facing fewer constraints than ND, as the crisis was Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 35 in its beginning. This implies that PASOK had ‘better’ overall economic conditions, but mainly had more control over policymaking during the first seven months of its governance, since it was only after May 2010 that Greece was submitted to the ‘Troika’ and the austerity policies, and the political implications of the crisis had not unfolded fully yet. All in all, the conclusions by Thomson et al. (2017) that were tested in the analysis found partial applicability and partial rejection. Findings that were supported in the crisis context relate to power-sharing arrangements, to the impact of institutional constraints, as well as to PASOK’s results. Findings on status quo pledges and ND’s separate result oppose Thomson’s et al. (2017) conclusions. The above discussion touched upon the specific objectives of this study. This thesis’ general objectives were to assess the bearing of normative expectations of the mandate theory of democracy (e.g. Downs, 1957) on contexts that do not facilitate the ‘normal functioning of democratic representation. The Greek cases, while provide conflicting evidence, further offer some useful insights. The policies implemented by PASOK after the submission to the ‘Troika’ were contradictory to many of its specific promises, and mainly to the overall ‘output’ of its program. In many cases, the austerity measures voted by the Parliament ‘cancelled out’ previously passed legislation; for example, the promise to ‘reduce the tax burden to the advantage of the low and middle classes’, was put forward by law 3842/2010, but was practically cancelled because of all the austerity measures that included horizontal increases in taxes. Put more generally, the Greek governments from 2010 and onwards, parallelly to their manifesto promises, have put forward policies and reforms that were not included in their electoral appeals. This is a very striking example of the limitations of the mandate model, which expects governing parties’ policies to be previously and clearly stated in parties’ election commitments. Finally, this study attempted to contribute to debates on governing parties’ policymaking ability in periods of crisis, with a special focus on Southern European democracies during the Great Recession. This endeavor has provided important insights on tribulations of governance under destabilizing conditions, and is a first attempt to cover a very important gap in the literature. Through my project, I have been able to highlight ways in which political trajectories might deviate from the expected direction if certain conditions are not met. Yet, much work is further needed in order to incorporate more examples and cases. This study contributes to this direction, but has several limitations that do not allow for generalizations or firm conclusions, especially regarding claims of causality. These limitations are discussed below. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 36 5.1. Limitations This study presents several limitations that need to be addressed. First, the comparative analysis is done solely qualitatively. This approach, although offers some important and initial understanding, it does not have explanatory in-depth power. The overall results show Greek pledge fulfillment under the period of pressure, however the findings do not imply that it was because of the pressure, therefore we should be raising awareness to draw causal inferences, since no such analysis was done. Ideally, a quantitative statistical analysis should have been conducted; that way, the findings would be statistically compared to those of Thomson et al. (2017), and other factors that explain, determine, alter, or condition the results would be added and controlled for. The findings on power-sharing for example, although in line with the compared results, do not give any information on whether it was indeed the constraints of sharing power that determined the low fulfillment. On a similar vein, the components of the ‘crisis’ have been taken into account as a whole, and they were not possible to be disentangled so as to rule out their independent effect on pledge fulfillment. My results therefore provide preliminary support for the hypotheses, and are restricted to descriptive evidences. These limitations indicate that my results should not be used as evidence of crisis effects. Second, the content analysis and the coding of the manifestos was done solely by me, and this raises some important concerns. Since no reliability tests with other native speakers were conducted, one should be careful regarding the validity and the reliability of the coding. The language limitations, as well as the limited scope of this project, did not allow for such tests, which should be undertaken in a future continuation of this project. However, my dataset along with documentation of the fulfillment judgements are available on request. Third, the pledges examined in the analysis are not entirely representative in that they include only PASOK’s most important promises, and only ND’s from the coalition government. This implies that not only do we not have the complete picture, but that the actual findings might eventually differ. Ideally, the analysis should include the total amount of PASOK’s promises, should integrate the election manifestos of PASOK and DIMAR for the 2012 elections, as well as the joint programmatic declaration of the tripartite coalition that was announced after the elections. In addition, the Greek pledges could have been coded according to their agreement or disagreement with the MoUs, and in that way the institutional impact of the ‘Troika’ on parties’ fulfillment could have been assessed better. Last, the four components of the pressure should be disentangled not only in order to examine their independent effect, but also to test for interactions effects, like for example between the economic crisis and the austerity policies, or the ‘Troika’ and the austerity measures. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 37 Apart from the above suggestions, further studies in Greek pledge fulfillment should focus on previous periods before the crisis. We appear to lack of empirical evidence on ND’s performance in pre-crisis times, therefore the extent to which ND was previously able to fulfill its promises remains unknown. Another interesting perspective to the pledge offer by Greek cases is the January-September 2015 SYRIZA coalition government. From a different angle, this case is of particular importance to the broader theoretical field of democratic mandates, or their breaking thereof. The failure of the SYRIZA-led government to implement its electoral program led to the interruption of its governing term, which associates the fulfillment of election promises with theories of government capacity and duration. Moreover, the impact of the ‘Troika’ in SYRIZA’s case was conclusive, which raises additional questions regarding the legitimacy of the democratic procedures when such institutions are so closely involved in national politics. All in all, Greece offers a very interesting and fertile ground for testing normative theories and empirical suggestions of democratic representation. 6. Conclusion The aim of this thesis was to examine the ability of incumbent parties to fulfill their election promises in times of crisis and under conditions of advanced pressure. The analysis focused on the increased strains Southern European democracies faced during the Great Recession, and particularly in Greece, the country mostly affected by the recent crisis. By examining the fulfillment of PASOK’s and ND’s promises prior to the 2009 and June 2012 elections respectively, I attempted to evaluate the program-to-policy linkage under such conditions. I further juxtaposed my findings derived from the crisis setting to results from ‘stable’ contexts and compared their applicability. Theoretically, I made use of a set of different pressures representative democracies under crises face, that could potentially restrict governing parties’ pledge fulfillment ability. This resulted in a theoretical framework that included the economic crisis, the austerity reforms, the limited maneuver space due to the presence of the ‘Troika’, and the political ramifications of the crisis that broadly pertained to a crisis of governance. Based on these, my main hypothesis tested the bearing of the general conclusion by Thomson et al. (2017), and suggested that Greek pledge fulfillment under the crisis would be lower compared to times and contexts of stability. The second hypotheses set out to test whether the additional obstacles coalition governments under crises face result in lower pledge fulfillment like under stable times, or whether the unification effect ostensibly established during the crisis could Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 38 nullify the additional constraints of power sharing. My third hypotheses examined two opposing scenarios regarding the fate of the status quo pledges in comparison to pledges aiming at change during times of pressure. Last, the fourth expectation suggested that ND’s promises that required negotiation with Greece’s lenders would have fewer prospects of being fulfilled. The results provided partial support for the first hypothesis, as the overall result of at-least partial fulfillment for both parties was 70 percent, and PASOK’s findings were even higher. ND’s percentage on the other hand was in line with that expected. In addition, the findings show support for the expectation that the coalition ND government would exhibit lower fulfillment rates than PASOK, and also strong evidence on the difficulty to maintain the status quo in crisis situations compared to changing it. Last, the fourth hypothesis was also corroborated as ND’s promises were significantly dependent on the bargaining with the ‘Troika’. These findings, although do not offer clear support for the hypothesized overall result, they nonetheless have valuable relevancy to modern politics and theories of democratic procedures. Yet, the important limitations of this study prevent us from gaining explanations and drawing conclusions on crisis effects on pledge fulfillment, as well as from generalizing the findings. In this project I attempted to offer some initial insights on democratic processes during periods of uncertainty. Despite its preliminary and exploratory character, this project showed that the normative expectations of traditional democratic theories, as well as the empirical evidence from ‘stable’ democracies might not be applicable to all conditions. It is evident that further research is required within contexts of crisis, since there is an important theoretical gap in this area that exceeds the strict time that confines of this project. The following research will look at how pledge fulfilment is affected in countries under financial and political pressure by doing cross comparisons across a variety of countries and parties, given that crises are more common than they used to be. In addition, future research should include more crisis situations and events, apart from economic crises, that would give prominence to a uniformity of crisis patterns and outcomes, under a common theoretical framework that can provide insights on political and governing turbulence. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 39 7. References APSA (American Political Science Association) (1950). Toward a More Responsible Two- Party System: A Report of the Committee on Political Parties. American Political Science Review, 44: Supplement Armingeon, K., & Baccaro, L. (2012). The Sorrows of Young Euro: The Sovereign Debt Crises of Ireland and Southern Europe. In N. Bermeo & J. Pontusson (Eds.), Coping with crisis: Government reactions to the great recession (pp. 162-197). New York: Russell Sage Foundation. Artés, J. (2013). Do Spanish parties keep their election promises?. Party Politics. Artés, J., & Bustos, A. (2008). Electoral promises and minority governments: An empirical study. European Journal of Political Research, 47, 307-33. Bartels, L. M. (2014). Ideology and retrospection in electoral responses to the Great Recession. In L.M. Bartels & N. Bermeo (Eds.), Mass politics in tough times: Opinions, Votes and Protest in the Great Recession (pp. 185-223). Oxford: Oxford University Press. Bartels, L., & Bermeo, N. (Eds.) (2014). Mass politics in tough times: Opinions, Votes and Protest in the Great Recession. Oxford: Oxford University Press. Baumgartner, F., & Jones, B. D. (1993). Agendas and Instability in American Politics. Chicago: The University of Chicago Press. Baumgartner, F., & Jones, B. D. (Eds.) (2002). Policy Dynamics. Chicago: The University of Chicago Press. Bermeo, N., & Pontusson, J. (2012). Coping with Crisis: An Introduction. In N. Bermeo & J. Pontusson (Eds.), Coping with crisis: Government reactions to the great recession (pp. 1- 31). New York: Russell Sage Foundation. Blyth, M. (2013). Austerity: The history of a dangerous idea. New York: Oxford University Press. Bosco, A., & Verney, S. (2012). Electoral epidemic: the political cost of economic crisis in Southern Europe, 2010–11. South European Society and Politics, 17(2), 129–154. Bueno de Mesquita, B. (1981). The War Trap. New Haven, CT: Yale University Press. Cameron, D. R. (2012). European Fiscal Responses to the Great Recession. In N. Bermeo & J. Pontusson (Eds.), Coping with crisis: Government reactions to the great recession (pp. 91-129). New York: Russell Sage Foundation. Chowanietz, C. (2011). Rallying Around the Flag or Railing Against the Government? Political Parties’ Reactions to Terrorist Acts. Party Politics, 17(5), 673–98. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 40 Costello, R., & Thomson, R. (2008). Election Pledges and their Enactment in Coalition Governments: A Comparative Analysis of Ireland. Journal of Elections, Public Opinion and Policy, 18 (3), 239-56. Cunha, L. (2008). Is the Dutch disease pandemic in the south? In F. Franco (Ed.), Challenges ahead for the Portuguese economy, 147–158. Lisbon, Portugal: Imprensa de Ciências Sociais. De Sousa, L., Magalhães, P. C., & Amaral, L. (2014). Sovereign Debt and Governance Failures: Portuguese Democracy and the Financial Crisis. American Behavioral Scientist, 58(12), 1517-1541. Diavgeia (2017). Ministry of the Interior and Administrative Reconstruction: Governmental Program Diavgeia [Υπουργείο Διοικητικής Ανασυγκρότησης: Πρόγραμμα Διαύγεια]. Available at https://diavgeia.gov.gr/search Downs, A. (1957). An Economic Theory of Democracy. New York: Harper. European Commission (2013). The Second Economic Adjustment Package for Greece: Third Review July 2013. Directorate-General for Economic and Financial Affairs. Occasional paper 159. Available at http://ec.europa.eu/economy_finance/publications/occasional_paper/2013/pdf/ocp159_e n.pdf Featherstone, K. (2015). External conditionality and the debt crisis: The ‘Troika’ and public administration reform in Greece. Journal of European Public Policy, 22(3), 295-314. Freire, A., Lisi, M., Andreadis, I., & Leite Viegas, J. (2014). Political Representation in Bailed- out Southern Europe: Greece and Portugal Compared. South European Society and Politics, 1-21. Gemenis, K., & Nezi, R. (2015). Government–Opposition Dynamics during the Economic Crisis in Greece. The Journal of Legislative Studies, 21(1), 14-34. Gladwell, M. (2001). The tipping point: How little things can make a big difference. London: Abacus. Government Gazette (2017). Ethniko Typografeio: Search: FEK [Εθνικό Τυπογραφείο: Αναζήτηση: ΦΕΚ]. Available at http://www.et.gr/index.php/anazitisi-fek Hays, J. C. (2003). Globalization and Capital Taxation in Consensus and Majoritarian Democracies. World Politics, 56(1), 79–113. Hellenic Parliament (2017). Legislative Work. [Νομοθετικό Έργο]. Available at http://www.hellenicparliament.gr/en/Nomothetiko-Ergo/Anazitisi-Nomothetikou-Ergou Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 41 Hellenic Statistical Authority (2017). [Ελληνική Στατιστική Αρχή]. Available at http://www.statistics.gr/en/home/ Hellwig, T. T. (2001). Interdependence, Government Constraints, and Economic Voting. Journal of Politics, 63(4), 1141–62. Hernandez, E., & Kriesi, H. (2016). The electoral consequences of the financial and economic crisis in Europe. European Journal of Political Research, 55(2), 203-224. Huber, E. & Stephens, J. D. (2001). Development and Crisis of the Welfare State: Parties and Politics in Global Markets. Chicago: Chicago University Press. Indridason, I. H. (2008). Does Terrorism Influence Domestic Politics? Coalition Formation and Terrorist Incidents. Journal of Peace Research, 45(2), 241–59. Kalogeropoulou, E. (1989). Election promises and governmental performances in Greece: PASOK’s fulfillment of its 1981 election pledges. European Journal of Political Research, 17, 289-311. Klingemann, H., Hofferbert, R., & Budge, I. (1994). Parties, policies, and democracy. Boulder: Westview Press. Kosmidis, S. (2014). Government Constraints and Accountability: Economic Voting in Greece Before and During the IMF Intervention. West European Politics, 37(5), 1-20. Kostadinova, P. (2013). Democratic performance in post-communist Bulgaria: election pledges and levels of fulfillment, 1997–2005. East European Politics, 29(2), 190-207. Kriesi, H. (2012). The Political Consequences of the Financial and Economic Crisis in Europe: Electoral Punishment and Popular Protest. Swiss Political Science Review, 18(4), 518- 522. Kriesi, H. (2014). The political consequences of the economic crisis in Europe: Electoral punishment and popular protest. In L.M. Bartels & N. Bermeo (Eds.), Mass politics in tough times: Opinions, Votes and Protest in the Great Recession (pp. 297-333). Oxford: Oxford University Press. Kriesi, H. (2015). Political Mobilization in Times of Crises: The Relationship between Economic and Political Crises. In M. Giugni & M. Grasso (Eds.), Austerity and Protest: Popular Contention in Times of Economic Crisis (pp. 19-33). London, UK: Routledge. Krippendorff, K. (2013). Content analysis: An introduction to its methodology (3rd ed.). Thousand Oaks, Calif.; London: SAGE. Lehmann, P., Lewandowski, J., Matthieß, T., Merz, N., Regel, S., & Werner, A. (2017). Manifesto Corpus. Version: 2017-2. Berlin: WZB Berlin Social Science Center. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 42 Lijphart, A. (1996). The puzzle of Indian democracy: A consociational interpretation. American Political Science Review, 90(2), 258-268. Lijphart, A. (1999). Patterns of Democracy: Government Forms and Performance in Thirty- Six Countries. New Haven, CT: Yale University Press. Lobo, M. C., & Lewis-Beck, M. S. (2012). The Integration Hypothesis: How the European Union Shapes Economic Voting. Electoral Studies, 31(3), 522–28. Special Symposium: Economic Crisis and Elections: The European Periphery. Lupu, N. (2013). Party Brands and Partisanship: Theory with Evidence from a Survey Experiment in Argentina. American Journal of Political Science, 57(1), 49-64. Mainwaring, S., Bejarano, A. M., & Lengómez, E. P. (2006). The crisis of democratic representation in the Andes: an overview. In S. Mainwaring, A. M. Bejarano, & E. P. Lengómez (Eds.) The crisis of democratic representation in the Andes (pp. 1-44). Stanford: Stanford University Press. Mair, P. (2006). Ruling the void?: The hollowing of Western democracy. New Left Review, (42), 25. Mair, P. (2009). Representative Versus Responsible Government. Max - Planck - Institut Für Gesellschaftsforschung. Working Paper 09 /8. Mair, P. (2011). Smaghi vs. the parties: Representative government and institutional constraints. Robert Schuman Centre for Advanced Studies and EU Democracy Observatory, EUI working paper no. 2011/22. Firenze: European University Institute. Mansbridge, J. (2003). Rethinking Representation. American Political Science Review, 97, 515- 28. Mansergh, L., & Thomson, R. (2007). Election pledges, party competition, and policymaking. Comparative Politics, 39(3), 311-330. Moury, C., & Freire, A. (2013). Austerity Policies and Politics: The Case of Portugal. Pôle Sud- Revue de Science Politique, 39(2), 35-56. Mudde, C. (2007). Populist Radical Right Parties in Europe. New York, NY: Cambridge University Press. Murphy, G. (2011). The Background to the Election. In M. Gallagher & M. Marsh (Eds.), How Ireland Voted 2011: The Full Story of Ireland’s Earthquake Election (pp. 1-29). Houndmills: Palgrave MacMillan. Naurin, E. (2009). Promising Democracy. Parties, citizens, and election promises. (Doctoral dissertation). Gothenburg studies in politics, 118. Gothenburg: Department of Political Science, University of Gothenburg. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 43 Naurin, E. (2011). Election promises, party behaviour and voter perceptions. Basingstoke: Palgrave Macmillan. Naurin, E. (2014). Is a Promise a Promise? Election Pledge Fulfilment in Comparative Perspective Using Sweden as an Example. West European Politics, 37(5), 1046-1064. Naurin, E., Håkansson, N., & Werner, B. (2014). Do Swedish Parties Fulfill Their Election Promises? Technical Report from the Project ‘If Elected, I Promise…’ Financed by the Swedish Research Council. Version April 2015. Department of Political Science, University of Gothenburg. Nezi, R., & Katsanidou, A. (2014). From valence to position: Economic voting in extraordinary conditions. Acta Politica, 49(4), 413. Nicolacopoulos, E. (2005). Elections and Voters, 1974–2004: Old Cleavages and New Issues. West European Politics, 28(2), 260-278. Pappas, T. (2003). The transformation of the Greek party system since 1951. West European Politics, 26(2), 90-114. PASOK (2009). Πρώτα ο πολίτης [The citizen (comes) first]. Πλαίσιο Κυβερνητικού Προγράμματος. Συνοπτική Παρουσίαση, Οκτώβριος 2009. Pierce, R. (1999). Mass-elite issue linkages and the responsible party model. In W. Miller, R. Pierce, J. Thomassen, R. Herrera, S. Holmberg, P. Esaiasson & B. Wessels (Eds.), Policy Representation in Western Democracies (pp. 9–32). Oxford: Oxford University Press. Polanyi, K. (1957). The great transformation. The political and economic origins of our time. Boston: Beacon Press. Pomper, G. M. (1968). Elections in America: Control and Influence in Democratic Politics. New York: Dodd, Mead and Company. Pomper, G. M., & Lederman, S.S. (1980). Elections in America: Control and Influence in Democratic Politics, 2nd ed. New York: Longman. Powell, B. G., & Whitten, G. (1993). A Cross-National Analysis of Economic Voting: Taking Account of the Political Context. American Journal of Political Science, 37(2), 391–414. Rallings, C. (1987). The influence of election programs: Britain and Canada 1945- 1979. In I. Budge, D. Robertson & D. Hearl (Eds.), Ideology, Strategy and Party Change: Spatial Analysis of Post-War Election Programs in Nineteen Democracies (pp. 1-15). Cambridge: Cambridge University Press. Remmer, K. (1991). The Political Impact of Economic Crisis in Latin America in the 1980s. The American Political Science Review, 85(3), 777-800. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 44 Riffe, D., Lacy, S., & Fico, F. (2005). Analyzing media messages: Using quantitative content analysis in research (2nd ed., LEA's communication series). London: Lawrence Erlbaum. Riker, W. H. (1964). Federalism: Origin, Operation, Significance. Boston, MA: Little, Brown. Roberts, K. M. (2013). Market Reform, Programmatic (De)alignment, and Party System Stability in Latin America. Comparative Political Studies, 46(11), 1422-1452. Rose, R. (1980). Do Parties Make a Difference? London: Macmillan. Royed, T. J. (1996). Testing the Mandate Model in Britain and the United States: Evidence from the Reagan and Thatcher Eras. British Journal of Political Science, 26, 45-80. Rüdig, W., & Karyotis, G. (2014). Who Protests in Greece? Mass Opposition to Austerity. British Journal of Political Science, 44(3), 487–513. Ryu, S.-R. (2009). Democratic Representation in New Democracies: Policy Responsiveness and Government Accountability (Doctoral dissertation). Retrieved from ProQuest Dissertations Publishing. (3372106) Sartori, G. (1976) Parties and Party Systems: A Framework for Analysis. Cambridge: Cambridge University Press Sotiropoulos, D. A. (2012). A Democracy under Stress Greece since 2010. Taiwan Journal of Democracy, 8(1), 27-49. Stiglitz, J. E. (2010). Freefall: America, Free Markets, and the Sinking of the World Economy. New York: W. W. Norton. Stokes, S. (2001). Mandates and democracy: Neoliberalism by surprise in Latin America. Cambridge: Cambridge University Press. Teperoglou, E., Freire, A., Andreadis, I., & Leite Viegas, J. (2014). Elites' and Voters' Attitudes towards Austerity Policies and their Consequences in Greece and Portugal. South European Society and Politics, 1-20. Thomson, R. (2001). The programme to policy linkage: The fulfilment of election pledge on socio-economic policy in the Netherlands, 1986-1998. European Journal of Political Research, 40, 171-197. Thomson, R., Royed, T., Naurin, E., Artés, J., Costello, R., Ennser-Jedenastik, L., Ferguson, M., Kostadinova, P., Moury, C., Pétry, F., & Praprotnik, K. (2017). The Fulfillment of Parties’ Election Pledges: A Comparative Study on the Impact of Power Sharing. American Journal of Political Science. doi:10.1111/ajps.12313 Tsatsanis, E., & Teperoglou, E. (2016). Realignment under Stress: The July 2015 Referendum and the September Parliamentary Election in Greece. South European Society and Politics, 21(4), 427-450. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 45 Tsirbas, Y. (2015). The January 2015 Parliamentary Election in Greece: Government Change, Partial Punishment and Hesitant Stabilisation. South European Society and Politics, 1-20. Vasilopoulou, S., & Halikiopoulou, D. (2013). In the Shadow of Grexit: The Greek Election of 17 June 2012. South European Society and Politics, 18(4), 523-542. Voulgaris, Y., & Nicolacopoulos, I. (2014). Εισαγωγή: O Eκλογικός Σεισμός του 2012. [Introduction: The Electoral Earthquake of 2012]. In Y. Voulgaris & I. Nicolacopoulos (Eds.), 2012: Ο Διπλός Εκλογικός Σεισμός [2012: The Twin Earthquake Elections] (pp. 9-31). Athens: Themelio. Walby, S. (2015). Crisis. Cambridge: Polity Press. Weschle, S. (2017). The Impact of Economic Crises on Political Representation in Public Communication: Evidence from the Eurozone. British Journal of Political Science, 1-20. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 46 Appendix 1. Coding of variables Below is the coding for all variables. Issue area and change-status quo are in more detail in Appendix 2. The coding of all variables is following Naurin, Håkansson, & Werner’s (2014) technical report. Coding for party is mine; the variable and the coding for ‘troika’ are also my additions. A. Party: PASOK equals 1, ND equals 2. B. Specific/Vague: Specific=1, Vague=2. C. Action/Outcome: Action=1, Outcome=2. D. Issue Area: see Appendix 2. E. Change/status quo: see Appendix 2. F. ‘Troika’ (only for ND’s promises): Negotiation=1, Without negotiation=2. G. Fulfillment: Fulfilled=1, Partially fulfilled=2, not fulfilled=3. PASOK’s pledges that were promised to be fulfilled within the first ‘100 days’ are coded as 1 if they fall within this threshold, 2 if they were fulfilled after the 100 days but within the governing period, and 3 if they were broken. Also, if these pledges were both fulfilled after this period, but also partially fulfilled in that the result was not fully what promised, are also coded as 2. The criterion according to which these pledges fall into the ‘100 days’ is the date the legislation was voted in the Parliament. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 47 Appendix 2. Coding schemes for issue area and change/status quo Table 7. Coding scheme and principles: issue/policy areas Code Issue area Examples Also 1 Agriculture food production issues dealt with under the ministry of agriculture farmers’ taxation/finances 2 Economics national finances taxation debt / public spending wages standard of living 3 Enterprise conditions for private enterprise ownership (private/public) privatizations competition monopoly 4 Employment, labour market (un)employment work environment labor legislation 5 Social welfare healthcare; childcare social insurance pensions pre school/kindergarden 6 Migration immigration integration of immigrants refugee issues immigration legislation 7 Education schools all levels (not pre school, see Social welfare). higher education research policy Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 48 8 Infrastructure roads, railroads, telecommunications ICT 9 Legal matters law enforcement police criminal law judiciary fire Department 10 European Union Greek membership Greek policy related to EU EMU/Euro 11 Environment ‘’Green Growth’’ Environmental law Renewable Energy Sources Protected areas; legislation Prevention of fires 12 Culture museums, theatre, music, etc. cultural activities/festivals archaeological sites tourism/islands sports Olympic Games 2004 facilities media policies, public media 13 Foreign policy defence, arm sales army – special forces borders issues under the Ministry of Exterior Greek-Turkish relations Greeks living abroad; absentee voting 14* Other political strategy issues (collaboration, elections etc) housing, planning 15 Political system/ public administration administrative reforms electoral laws constitution ‘Kallikratis’ program Public services Note: Table and overall principles taken by the technical report (version April 2015) for Naurin, Håkansson, & Werner (2014). Category 14 was not used in my coding. Category 15 is my addition. Examples taken from my data, and differentiating principles applied in my coding are highlighted with bold letters. The rest are the same. Under pressure: Democratic Mandates and Policy Outcomes in turbulent times. Evidence from Greece 49 Table 8. Coding scheme and principles: Change/status quo Code Label Definition Examples 1 Status quo pledges to maintain present conditions “No new tax burden in middle and low incomes” 21 Change: cut pledges to reduce public spending in any area, to decrease or abolish publicly financed undertakings “reduced benefits for those with higher incomes are necessary” 22 Change: expense pledges to increase public spending, to increase or initiate publicly financed undertakings “Increases in salaries above inflation” 'Extension of unemployment benefit from 1 to 2 years'' 23 Change: tax cuts pledges to cut or abolish taxes and/or dues 'Reduction of taxation so that the tax burden doesn't exceed the 25% of the average family income'' 24 Change: tax increase pledges to raise or initiate taxes and/or dues “we will raise the income tax” Also, indirectly: ''Abolition of a series of gratuitous tax exemptions established the past 5 years'' 25 Change: other all change pledges that do not fall into 21- 24 above 'Strengthening and upgrading the role of the Hellenic Competition Commission'' 3 Review Pledges in which the party suggests a review/inquiry/investigation of an issue, promising neither substantial change, nor status quo 'Reexamination of the institutional framework for dealing with noise pollution'' Note: Table and overall principles taken by the technical report (version April 2015) for Naurin, Håkansson, & Werner (2014). Examples taken from my data are highlighted with bold letters.