Party Politics http://ppq.sagepub.com/ Economy, corruption or floating voters? Explaining the breakthroughs of anti-establishment reform parties in eastern Europe Sean Hanley and Allan Sikk Party Politics published online 18 September 2014 DOI: 10.1177/1354068814550438 The online version of this article can be found at: http://ppq.sagepub.com/content/early/2014/09/17/1354068814550438 Published by: $)SAGE http://www.sagepublications.com On behalf of: Political Organizations and Parties Section of the American Political Science Association Additional services and information for Party Politics can be found at: Email Alerts: http://ppq.sagepub.com/cgi/alerts Subscriptions: http://ppq.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav » OnlineFirst Version of Record - Sep 18, 2014 What is This? Downloaded from ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 Article PařfyPolitics Party Politics Economy, corruption or floating Author ml Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1354068814550438 ppq.sagepub.com voters? Explaining the breakthroughs of anti-establishment reform parties pp®sage in eastern Europe Sean Hanley UCL School of Slavonic and East European Studies, London, UK Allan Sikk UCL School of Slavonic and East European Studies, London, UK Abstract This paper discusses a new group of parties that we term anti-establishment reform parties (AERPs), which combine moderate social and economic policies with anti-establishment appeals and a desire to change the way politics is conducted. We analyse the electoral breakthroughs of AERPs in central and eastern Europe (CEE), the region where AERPs have so far been most successful. Examples include the Simeon II National Movement, Movement for the European Development of Bulgaria (GERB) (Bulgaria), Res Publica (Estonia), New Era (Latvia), TOP09 and Public Affairs (Czech Republic) and Positive Slovenia. We examine the conditions under which such parties broke through in nine CEE states in 1997-2012 using fuzzy-set qualitative comparative analysis (fsQCA). We find five sufficient causal paths combining high or rising corruption, rising unemployment and party system instability. Rising corruption plays a key role in most pathways but, unexpectedly, AERP breakthroughs are more closely associated with economic good times than bad. Keywords New parties, QCA, party system change, corruption, eastern Europe Introduction Mobilization against Europe's political establishments is on the rise. This has naturally focused the attention of scholars and policy-makers on the emergence of protest parties and movements. Much discussion has centred on the electoral successes and prospects of populist radical right parties such as Hungary's Jobbik or the True Finns and groupings on the radical left like SYRIZA in Greece (Jordan, 2010; Wolin, 2011). Fears of a radical populist electoral backlash have been especially marked in relation to central and eastern Europe (CEE), whose weaker economies and less consolidated democracies appear fertile ground for radical right and illiberal populist parties after the falling away of the European Union (EU) accession conditionalities (Bohle and Greskovits, 2009; Rupnik, 2007). However, alongside conventional radical populists a major new protest phenomenon has appeared which so far has been relatively under-researched: successful new parties, which combine mainstream ideology on economic and socio-cultural issues with fierce anti-establishment rhetoric and demands for political reform, transparency and new ways of 'doing polities'. Such parties matter politically. They can achieve overnight electoral breakthroughs on a scale sufficient to restructure party systems and unlike radical populist groupings, even when they do not, often have high coalition potential. Such protest-oriented parties may pose a challenge for democracy. Despite their success, they often struggle to govern and can rapidly break up, sometimes preparing the ground for new protest parties, Paper submitted 23 June 2014; accepted for publication 5 July 2014 Corresponding author: Sean Hanley, UCL School of Slavonic and East European Studies, Gower Street, London WCIE 6BT, UK. Email: s.hanley@ucl.ac.uk Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 2 Party Politics potentially feeding a spiral of protest and instability (Deegan-Krause, 2007; Deegan-Krause and Haughton, 2009). New anti-establishment parties of this kind are increasingly breaking through in established democracies, Beppe Grillo's Five Star Movement being perhaps the most spectacular example. However, they have been a feature across the political landscape of CEE for some years (Mesežnikov et al., 2013). Parties such as the Simeon II National Movement in Bulgaria founded by the notional heir to the Bulgarian throne (Barany, 2002) or Res Publica in Estonia (Taagepera, 2006) enjoyed electoral landslides months after launching in 2001 and 2003, respectively, and immediately became leading parties in the government. Others, such as Freedom and Solidarity (SaS) in Slovakia (2010) or Public Affairs (VV) (2010) and the Akce nespokojených občanů (ANO) movement of billionaire Andrej Babiš (2013) in the Czech Republic have achieved more modest success, but entered government coalitions. Elsewhere in CEE, however, such parties have been conspicuous by their marginality or absence. The experience of CEE thus represents a natural laboratory for the comparative study of such parties, which casts an important light on the wider prospects of this emerging type of protest party. We term such parties anti-establishment reform parties (AERPs) and in this article we analyse the conditions under which their electoral breakthroughs occur using fuzzy-set qualitative comparative analysis (QCA) (fsQCA). We first discuss our conceptualization of the AERP, briefly relating it to existing concepts such as populism, and discuss parties in CEE that we classify as AERPs. We then present the fsQCA method and briefly review its use in analysing the comparative success of emergent parties across different national and electoral contexts. We then consider a range of causal conditions relevant to AERPs' success including economic crisis, corruption and party system instability and, using fsQCA, pick out those configurations leading to AERP breakthrough. Contrary to our initial expectations we find that breakthroughs can occur during either good or bad economic times and that it is a rising perceived corruption which is common to many economic and political contexts leading to AERP breakthrough. AERPs As Deegan-Krause (2010) observes, despite much diversity, there are clear commonalities between many new parties that have broken through in recent elections in CEE making them '... not exactly a new party family (though in their cultural liberalism and anti-corruption emphases they share significant elements) and not exactly a new party type ... but with strong and intersecting elements of both'. How can this phenomenon be conceptualized? Authors who have noticed the phenomenon have often viewed such parties as a sub-type of populism, speaking of 'new/centrist populism' (Pop-Eleches, 2010) and 'centrist populism' (Učeň, 2007; Učeň et al., 2005). Others see them more narrowly as based on a distinct issue dimension; Bágenholm (2013a) for example, terms them 'anti-corruption parties'. We conceptualize these parties somewhat differently as AERPs which exhibit—to different extents—three core features: (1) a politics of mainstream reformism (2) usually framed in terms of anti-establishment appeal to voters and (3) genuine organizational newness. By mainstream reformism we understand the following: firstly, that these parties are committed to mainstream models of liberal democracy and the market economy and display neither the populist radical right's inclination to 'illiberal democracy', ethnocentrism and social conservatism (Mudde, 2007) nor the anti-capitalism of the radical left (March and Mudde, 2005). Secondly, AERPs display a strong commitment to political reform, seeking to reform political institutions or change the way politics is conducted. Thus depending on context, they may make appeals to fight corruption; replace corrupt or inefficient elites; create new democratic structures; or simply offer novelty of political style—a 'project of newness' as Sikk (2012) terms it. Following Abedi (2004: 12) we further understand AERPs as anti-establishment parties: parties which present themselves as challengers to establishment parties and emphasize the divide between society and the political establishment. The third element of our definition is that organizationally AERPs are 'genuinely new'. Here we apply, in slightly extended form, Sikk's (2005: 399) definition of new parties as those that are 'not successor to any previous parliamentary parties, have a novel name and structure, and do not have any important figures from past democratic politics among their major members'. We thus exclude alliances and mergers between established parties and parties resulting from breakaways from established parties.1 Our stress the on anti-establishment stance of AERPs overlaps with the widely used concept of 'populism'. However, for a number of reasons we avoid this label. Even when clearly and minimally defined (see Mudde (2004: 23)) the concept tends to conflate anti-establishment appeal with a set of moralistic anti-political appeals. While often empirically associated, these are, we contend, conceptually distinct and do not logically imply one another, leading to empirical miscategorization. Although often labelled 'populist', few AERPs make the strongly moralistic distinctions between the political elite and the pure unsullied people said to be a defining characteristic of populist parties. We also take issue with the term's normative connotations (see Sikk (2010) for further discussion). We thus use the concept of AERP as a broad working category, whose validity will be tested by the search for common causal patterns. To identify AERPs empirically we first identify genuinely new parties and eliminate those regarded as radical right or radical left in the literature Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 Hartley and Sikk 3 (March, 2011; Mudde, 2007) or whose programmes or declarations clearly identify them as such. We then examined party programmes and statements and used case knowledge to distinguish those new parties making anti-establishment reformist appeals from other new parties. We thus excluded a number of new parties including radical right groupings (League of Polish Families (2001), Atoka, Bulgaria (2005) and Jobbik, Hungary (2010)) and green parties (Estonia (2007) and Czech Republic (2006)). Strikingly, however, the large majority of successful genuinely new parties in CEE have been AERPs with only a minority of successful new parties in the region emerging on the radical right. Although potentially limiting leverage for distinguishing the conditions of AERP success from those of other types of anti-establishment party, this does not reduce our study to a generic study of new parties. Although different types of successful new parties may share some causal drivers, AERPs are a conceptually distinct subset of new parties whose success is encompassed by a distinct range of causal conditions2 which we study through a cross-national, cross-election comparison of distinct pathways. We identified 21 successful AERPs in parliamentary elections in CEE between 1994 and 2012. In all instances, we refer to a party at the time of the parliamentary election in question as CEE parties, both new and established can experience considerable changes in their identity and programmatic appeals. Our data clearly suggests that AERP breakthrough is largely a phenomenon of approximately the last 15 years. This fits the observations of Pop-Eleches (2010: 223) regarding the timing of the success of 'unconventional parties' in CEE, which he explains in broad aggregate terms by the dynamics of 'third generation' post-communist elections. Having voted into office and become disillusioned with the performance of conventional parties of the left and right in successive elections, voters are ready to turn to unconventional new parties. We accept this logic and focus on 'third generation' elections to the lower houses of CEE parliaments, which took place between September 1997 and December 2012. We list the elections and AERPs covered in Table l.3 No AERPs have materialized in Romania which we contend may be related to the markedly low level of democratic freedoms in that country compared to other EU states (as indicated by Freedom House political rights and press freedom scores). For the sake of analytical clarity we therefore exclude this country from our analysis. QCA conditions For our analysis we use QCA, a comparative technique which formalizes the logic of qualitative case-based comparison by linking configurations of causes (conditions) to effects (outcomes) using Boolean algebra and set theory (Ragin, 2008; Rihoux and Ragin, 2009). QCA is well suited to cross-national comparison of the success levels of a group of new parties as it can address both relatively high numbers of cases and high levels of casual complexity, capturing common causes, configurations of causes and multiple pathways to the same outcome (Gherghina and Jiglau, 2011; Redding and Viterna, 1999; Veugelers and Magnan, 2005). We use fsQCA where cases are coded in terms of their degree of set membership in outcome and causal conditions, rather than the dichotomous presence or absence of conditions and outcomes as in the original crisp set version of QCA (csQCA, see Ragin (2008)). Degrees of membership are expressed as values ranging from 1.0 (full membership) to 0.0 (full non-membership) with a 'crossover value' of maximum ambiguity set at 0.5. Although expressed numerically, the degrees of set membership are anchored in researchers' theoretically-based judgments, with at least three key anchor points (0, 0.5 and 1) corresponding to a verbal description.4 Outcome: AERP electoral breakthrough (BREAKTHRU) AERPs have considerably greater vote winning potential than the niche or radical parties which earlier QCA studies of new party emergence focused on (Gherghina and Jiglau, 2011; Redding and Viterna, 1999; Veugelers and Magnan, 2005). In the elections we studied, there were two cases—Bulgaria in 2001 and 2009—where a single AERP was supported by more than a third of the electorate and one (Slovenia in 2011) where the combined vote for AERPs was above this level. We set the threshold for full membership in the set AERP electoral breakthrough (BREAKTHRU) at a level of massive electoral support (30% of votes or more) when the AERP becomes the first or second biggest party and hence a major party in a coalition government or a major opposition party.5 We set the crossover point (0.5 set membership) at 7% of the vote, which is safely sufficient to win parliamentary representation and to gain a share of seats roughly proportional to the party's vote share, with the AERP becoming a minor governing or opposition party. We deem any election where there is no AERP vote to be fully out of the set (0).6 Causal conditions Previous cross-national QCA studies of the comparative success of green, far-right and ethnic parties (Gherghina and Jiglau, 2011; Redding and Viterna, 1999; Veugelers and Magnan, 2005) have drawn on ready-made propositions about causes of their success from a large existing literature. For an emerging group of parties such as AERPs, we necessarily draw on a more diverse range of literature: the limited work on AERP-type parties (Bágenholm, 2013a; Sikk, 2012; Učeň et al., 2005); our own case knowledge of CEE; the comparative literatures on new parties, populism, voting behaviour and democratic quality; and recent commentary by academic writers. Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 4 Party Politics Table I. Electoral support for anti-establishment reform parties (AERPs) in central and eastern Europe 1997-2012. Election Successful AERP Votes % Set membership in BREAKTHRU BGR 2001 Simenon II Movement (NDSV) 42.7 1.00 BGR 2005 - 0.0 0.00 BGR 2009 Movement for the European Development of Bulgaria (GERB) 39.7 1.00 CZE 2002 - 0.0 0.00 CZE 2006 - 0.0 0.00 CZE 2010 TOP09, Public Affairs (VV) 27.6 1.00 EST 1999 - 0.0 0.00 EST 2003 Res Publica 24.6 0.99 EST 2007 - 0.0 0.00 EST 2011 - 0.0 0.00 HUN 1998 - 0.0 0.00 HUN 2002 - 0.0 0.00 HUN 2006 - 0.0 0.00 HUN 2010 Politics Can Be Different (LMP) 7.5 0.54 LTU 2000 New Union (SL) 19.6 0.98 LTU 2004 Labor Party (DP) 28.4 1.00 LTU 2008 National Resurrection Party (TPP) 15.1 0.92 LTU 2012 Way of Courage (DK) 9.8 0.70 LVA 1998 New Party (JP) 7.3 0.53 LVA 2002 New Era (|L) 24.0 0.99 LVA 2006 - 0.0 0.00 LVA 2010 - 0.0 0.00 LVA 201 1 Zatler's Reform Party (ZRP) 21.3 0.99 POL 1997 - 0.0 0.00 POL 2001 Law and Justice (PiS) 9.5 0.68 POL 2005 - 0.0 0.00 POL 2007 - 0.0 0.00 POL 201 1 Palikot Movement (RP) 10.5 0.74 SVK 2002 SMER, Alliance of the New Citizen (ANO) 21.5 0.99 SVK 2006 - 0.0 0.00 SVK 2010 Freedom and Solidarity (SaS) 12.2 0.82 SVK 2012 Ordinary People (OL'aNO) 1 1.4 0.79 SVN 2008 - 0.0 0.00 SVN 201 1 Positive Slovenia (PS-LZJ), Virant List (DLGV) 36.9 1.00 Source: European Elections Database and websites of national electoral authorities. Note: scores for parties with less than 4% were not used, resulting in formal raw score of 0.0. BGR - Bulgaria; CZE - Czech Republic; EST - Estonia; HUN -Hungary; LTU - Lithuania; LVA - Latvia; POL - Poland; SVK - Slovakia; SVN - Slovenia. Figures refer to year in which an election took place. Crisis and economic hard time. Many commentators have seen the recent rise of anti-establishment parties across Europe as a response to the pressures exerted by the global economic downturn and the Eurozone crisis (Bartels, 2013; Cramme, 2013; vanBiezen and Wallace, 2013: 294-7). Such expectations echo the political science literature which sees the inability of established actors to deliver reasonable economic conditions as a potential driver of protest voting for new parties. Such patterns of anti-incumbent economic voting generally are, moreover, sharper and more prevalent in CEE (Roberts, 2009). The global economic downturn that started in 2008-9 affected all CEE states following a period of economic expansion across the region. However, the downturn affected CEE states to markedly different degrees and built on pre-recession economies which also varied, particularly in terms of unemployment, reflecting different national trajectories of post-communist reform. We therefore look at two economic conditions expected to enhance the chances of AERP breakthrough: economic contraction (NOGRO) and rising unemployment (INCU-NEMP). Economic contraction (NOGRO) has both direct effects on consumption and levels of public provision and broader socio-political effects in stoking public discontent with incumbents. Extreme levels of economic contraction also create a sense of social crisis which may prompt voters to look for political alternatives. To operationalize economic contraction (NOGRO), we set the threshold for full set membership at a 5% decline in a country's gross domestic product (GDP). The minimum threshold ('fully out') is set at a very high level of economic growth of 5%, a phenomenon empirically Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 Hartley and Sikk 5 observed in CEE states at various points since 1989. We set the crossover point at 0% annual change in GDP, corresponding to a stagnant economy. We look at the average level of economic growth over two previous years, as the effects of growth might become visible with a slight delay. Likewise, we expect that a period of economic downturn, especially if marked, may have socio-psychological impact felt even after the economy has subsequently picked up. A second consequence of economic downturn is increasing unemployment. This directly and immediately impacts those made redundant and their families, but arguably also affects wider groups of voters who fear for their own job security.7 Even at relatively low levels, sharply increasing unemployment represents a favourable condition for a turn to non-establishment politics and AERP breakthrough. Hence, we incorporate a condition of sharply increasing unemployment (INCUNEMP) which we operationalize as the change in unemployment rates over the two years before the election. We set the maximum threshold corresponding to full membership in the set at an increase of three percentage points and the lower threshold ('fully out of set') at a three percentage point decrease in unemployment levels. The crossover point is set at a near zero decrease of 0.5 percentage points.8 These two economic conditions define four distinct socio-economic conjunctures characteristic of different countries and time periods: economic recession where GDP is falling and unemployment increasing (NOGRO*UNU-NEMP); economic boom where GDP is increasing and unemployment falling (~NOGRO* ~INUNEMP); restructuring or reform where the economy is growing but shedding jobs (~NOGRO*INCUNEMP); and recession with a social safety valve where unemployment falls despite economic contraction (NOGRO* ~ INCUNEMPs) spurred by mass withdrawal from the labour market due to emigration or policies promoting early retirement. Perceived corruption and distrust. In contrast to explanations which foreground economic recession and growth in unemployment, some authors have interpreted the rise of anti-establishment parties as a crisis of confidence in conventional democratic politics and the honesty and competence of elites (Kaldor and Selchow, 2013; Zizek, 2013). Perceived corruption and the politicization of corruption have often been linked to the rise of AERP-like parties (Bagenholm, 2013a; Deegan-Krause, 2010) and, more broadly, to anti-incumbent voting (Bagenholm, 2013b; Slomczynski and Shabad, 2012). Such anti-corruption sentiment may be understood both in terms of direct concern about corruption and an inchoate sense that political elites are self-serving, untrustworthy and unrepresentative. However, other writers (Blass et al., 2012; Hooghe and Quintelier, 2014) associate high levels of perceived corruption with demobilization and non-participation benefiting established politicians. For this reason we distinguish two corruption-related conditions: a social perception of high corruption (HICORR) and a substantial increase in perceived levels of corruption (INCORR). For the purposes of QCA analysis we initially assume both to have a positive effect on AERP breakthrough. To operationalize levels of perceived corruption we use Transparency International's annual corruption perception index (CPI). Although CPI is often criticized as poorly reflecting 'real' levels of corruption, it is arguably a meaningful measure of the type of public concerns we wish to highlight.9 We place an election fully in the set of high levels of perceived corruption (HICORR) if the CPI score for the election year10 falls below 3.5, around the worst empirically achieved levels in CEE after 1999 (Romania and Bulgaria). A case is fully out of this set if CPI reaches 5.5—a benchmark level based on the lowest levels of perceived corruption in the region after this date (achieved only by Slovenia and Estonia). The crossover point of maximum ambiguity (0.5) is 4.6, a figure close to the median corruption rating in CEE across the period. We deem a case to be fully in the set of elections where there has been a substantial increase in the level of perceived corruption (INCCORR) if the CPI score decreases by 0.4 points, indicating a substantial deterioration in corruption. A case is fully out of the condition if a country's CPI score increases by 0.4 points over the preceding two years—that is there is substantial improvement.11 We set the crossover point at a decrease in the CPI score of just over zero (0.01), a point where there is neither improvement nor deterioration.12 Political conditions. In earlier iterations of this work (Hanley and Sikk, 2011) we included several conditions relating to the party-electoral context, including the presence of strong radical (right) populist parties and electoral turnout. However, these lacked explanatory power. With increased turnout, it was difficult satisfactorily to distinguish cause and effect, while most electorally significant radical parties in CEE were, on closer inspection, relatively niche groupings without insufficiently broad appeal to compete with AERPs. In this paper we retain just one such condition: previous levels of voting for genuinely new parties (HGENP). CEE parties and party systems are weakly institutionalized compared to other regions. However, there have been uneven patterns of party system stability within the region (Powell and Tucker, 2014) which may influence AERP breakthrough. We hypothesize that a history of support for genuinely new parties reflects the presence of a significant pool of 'available' voters, who may perceive an emerging AERP as a credible challenger. To operationalize this condition we took the maximum support for genuinely new parties in the previous two elections. A case is a full member of this condition (set membership 1.0) if combined support for genuinely new Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 6 Party Politics parties was 30% or more—enough to generate one new major party or a number of minor breakthroughs. A case was fully out of this set if no genuinely new party won votes over this period. The crossover point was set at 19%, equating to substantial support for one genuinely new party or more modest support for a range of less successful new parties.13 Empirical analysis We used the QCA module in R (Du§a and Thiem, 2012), to analyse BREAKTHRU in relation to these five sociopolitical conditions. As expected we found that there was no single necessary condition for AERP breakthrough. This confirms that AERP breakthroughs cannot be accounted for by the all-encompassing narratives often used to frame them: popular reactions to economic hard times or a crisis of party systems and party government. The two single conditions which came closest to being necessary were rising corruption and rising unemployment. However, their consistency scores of 0.721 and 0.622 respectively placed them far below what is required for causal necessity.14 One combination of two conditions (HICORR + INCORR) had high enough consistency (0.91) to be regarded as a necessary condition. This states that AERP breakthrough required either rising corruption or an already high level of corruption. However, its coverage of cases (0.58) was relatively low. This suggests that accounts which stress the bubbling up of anti-political, anti-corruption sentiments may come closer to general explanation but are far from sufficient. Seeking sufficient paths to AERP breakthrough To find sufficient causal paths we generated a 'truth table' which shows the consistency of the 32 possible combinations of conditions relative to AERPs breakthrough (see the online supplementary materials15). Individual elections are listed in the rows (causal combinations) with which they are most consistent. Rows of logically possible counterfactual combinations of conditions with no matching real life case ('logic remainders') are also shown. To determine which causal configurations should be classified as leading to BREAKTHRU we set a consistency cut-off at 0.78. This is slightly below the widely used cut-off of 0.8, but comfortably above the 0.75 minimum recommended in the literature (Rihoux and Ragin, 2009: 87-112; Schneider and Wagemann, 2012: 279) and reflects a natural gap in the distribution. In line with normal fsQCA practice we first examined the conservative (complex) solution produced using only empirically-occurring cases. The conservative solution had a high level of consistency (0.85) and relatively broad coverage (0.79 and identified five relatively complex sufficient causal paths for AERP breakthrough. We then simplified our solution by incorporating counterfactual cases. We formulated a parsimonious solution including all such counterfactual 'logical remainders' and an intermediate solution incorporating only 'good counterfactuals' (Schneider and Wagemann, 2012: 168-175, 199) offering a middle way between parsimony and empirical complexity.16 The 'easy' or 'good' counterfactuals used in the intermediate solution require clearly stated, theoretically or empirically based assumptions about 'directionality': the causal effects that conditions would have in the counterfactual cases. QCA studies focusing on testing existing theory typically derive these off-the-shelf from large mature literatures. However, for a theory-building undertaking such as the comparative analysis of a new emerging group of parties, where the literature was inevitably more limited, we proceeded more cautiously. In setting the directional assumptions we drew on the limited published work on AERPs (largely single country case studies), our own case knowledge of the region, insights that can be gleaned from related literature on new parties or populism and careful consideration of patterns produced by the initial conservative solution. Reviewing our conditions, we concluded that four could plausibly be interpreted as contributing to AERP breakthroughs in either positive or negative form depending on the wider configuration of causes. Thus, while unstable party systems (HGENP) might (as widely argued in the literature) provide opportunities for new populist-type parties, stable party systems (~ HGENP) could do the same in some contexts if they had become rigid, unresponsive or oligarchical (Kaltwasser, 2012). Rising unemployment and economic contraction might, as initially anticipated, drive electoral discontent with establishment parties. However, examining key cases in the conservative solution (Estonia (2003), Lithuania (2004) and Lithuania (2008)) we judged that in some contexts falling unemployment and a buoyant economy might provide a cue for some voters to turn away from economic issues and focus on questions of corruption and governance—opening up opportunities for AERPs. As already noted, high corruption can favour new anti-establishment parties in many contexts and demobilize in others. Conversely, we concluded, there might also be contexts where low perceived corruption could also plausibly create circumstances favourable to AERP breakthrough. Low but rapidly increasing corruption (INCCORR * ~ HICORR) which appears in two of our paths (2, 4) might have an especially shocking and mobilizing effect. We therefore set a directional expectation only for rising corruption (INCORR), which theoretical and empirical evidence consistently pointed to as favouring AERP breakthrough. The intermediate solution (Table 2) enabled us to identify five distinct contexts favourable to AERP breakthrough: Path 1: corrupt socially painful growth (~NOGRO* INCUNEMP*HICORR). This scenario combines rising unemployment (INCUNEMP) with economic Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 Hartley and Sikk 7 Table 2. Intermediate solution (BREAKTHRU). Unique Consistency PRI Coverage coverage Cases 1 ~NOGRO*INCUNEMP* HICORR 0.905 0.890 0.370 0.204 BGR0I, BGR09, LTU00, POLO 1, SVK02, SVKI0, SVKI2 2 ~NOGRO*INCCORR* HGENP 0.841 0.810 0.351 0.105 BGR09, LTU04, LTU08, LTU12, LVA02, SVKI0, SVKI2 3 ~ NOGRO*-INCUNEMP* 0.889 0.870 0.157 0.053 EST03, LTU04, LTU 12 HICORR*INCCORR 4 NOGRO*INCUNEMP* 0.831 0.793 0.150 0.132 CZE 10, HUN 10, SVNI 1 INCCORR*-HGENP 5 NOGRO*HICORR* INCCORR*-HGENP 0.985 0.984 0.073 0.056 LVAI1 Note: consistency: 0.85, coverage: 0.81. BGR - Bulgaria; CZE - Czech Republic; EST - Estonia; HUN - Hungary; LTU - Lithuania; LVA - Latvia; POL - Poland; SVK - Slovakia; SVN - Slovenia. Figures refer to year in which an election took place. growth (~ NOGRO) and a background of high perceived corruption (HICORR). This corresponds to a context of apparently successful economic reform or restructuring, whose costs and benefits are, nevertheless, seen as unjustly distributed because of unemployment and high levels of perceived corruption. This experience was largely characteristic of a phase of post-communist reform in some CEE states shortly before EU accession in 2000-2 (Lithuania (2000), Poland (2001) and Slovakia (2002)) as well as of Slovakia in 2010 and 2012 as it recovered from the 2008-9 recession. Path 2: growth but increasing corruption in an unstable party system (~ NOGRO* INCCORR* HGENP). This path, which covers Lithuania (2004-2012), Latvia (2002), Slovakia (2010, 2012) and Bulgaria (2009) shows economic growth (~ NOGRO) coupled with increasing corruption (INCCORR) in an unstable party system (HGENP) favouring AERP breakthrough. The configuration suggests that even where the economy is growing, when corruption is increasing, voters will turn to AERPs in large numbers if there is already a tradition of voting for new parties. Path 3: low and rising corruption in economic good times (~ NOGRO* ~INCUNEMP* ~ HICORR * INCCORR). Path 3 covers three Baltic elections (Estonia (2003), Lithuania (2004) and Lithuania (2012)) and like path 1 highlights how corruption can interact with a seemingly benign socio-economic climate. In these cases a favourable context for AERP breakthrough is created by low but rising corruption (~HICORR*INCCORR) and a buoyant economy with both growth and falling unemployment (~NOGRO* ~ INCUNEMP). Increases in perceived corruption in a relatively low corruption environment we suggest have a galvanising effect, while improvement in the economy allows voters to refocus on issues of corruption and governance. Path 4: recession and rising corruption in rigid party systems (NOGRO*INCUNEMP*INCCORR* ~ HGENP). Path 4 is a distinct sub-regional path featuring only elections in three recession-hit Visegrád states with previously stable party systems—Hungary in 2010, the Czech Republic in 2010 and Slovenia in 2011. These had been generally resistant to AERP breakthroughs until the first elections after the 2008-9 downturn. At this point, a configuration of recession (NOGRO*IN-CUNEMP), rising perceived corruption (INCC ORR) and the previous stability of the established party system (~ HGENP) combined to create favourable conditions for AERP breakthrough. In all three cases the inability of (some or all) established parties credibly to respond to economic crisis and their loss of legitimacy because of growing concerns over corruption prepared the ground for AERP breakthrough. Strikingly, in this configuration party stability rather than party system fluidity contributes to AERP breakthrough: long established parties appeared ossified, corrupt, out-of-touch and an obstacle to both the solution of urgent socio-economic problems and longer term modernization (Batory, 2010; Haughton and Krašovec, 2011; Haughton et al., 2011). Path 5: Latvia's way? (NOGRO*HICORR*INC-CORR * ~ HGENP). The breakthrough of the Zatlers Reform Party in Latvia's 2011 election appears as a unique case represented by its own causal path, albeit in some ways one close to the recession and rising corruption in rigid party systems path experienced in Visegrád states (path 4). Here too economic contraction (NOGRO) combined with an increasingly stable, but oligarchical party establishment. However, the economic context was characterized by economic contraction without rising unemployment because of an unusually sharp and deep recession and high levels of emigration. Latvia's path to AERP Downloaded Irum ppq.sagepub.com at Masarykova Univerzita on September 19, 2014 8 Party Politics INTERMEDIATE SOLUTION q