From Electoral Epidemic to Government Epidemic: The Next Level of the Crisis in Southern Europe

Abstract

With the elections of 2015–16 in Greece, Spain and Portugal, the political fallout from the economic crisis in Southern Europe reached a new level. An overview of the new electoral arenas and party systems is followed by an investigation of the paths to government formation. Unprecedented events during (multiple) processes of government formation, uncharted outcomes in government types and governing party identities, and the necessity to repeat elections are three features defining a new syndrome that first appeared in Greece but has now spread to Iberia. In Southern Europe the road to incumbency has become arduous and the risk of government instability high.

KEYWORDS:

• Greece
• Spain
• Portugal
• critical elections
• party systems
• coalition government
• government formation
• incumbency

Inconclusive elections that had to be repeated; the breakdown of established patterns of political representation; a winning party that refused a mandate to form a government; a winning coalition rejected by a no confidence vote; a government coalition of the radical left and nationalist right; a left–left cabinet agreement; and a case of parties repeatedly failing to form any coalition government. These are just some of the unprecedented events that have recently marked the perilous path from elections to government installation in Southern Europe.

In the first phase of the economic crisis (2010–11), Southern Europe faced the outbreak of an ‘electoral epidemic’ whose hallmarks were declining voter turnout, severe incumbent punishment without special benefit for the official opposition, and the emergence of new contenders (Bosco & Verney 2012). A second phase in 2012–13 consolidated ‘protest elections and challenger parties’ as the prevalent characteristics of the new electoral scene (Verney & Bosco 2015). In the following years, the electoral epidemic has spread further. With the last wave of general elections in Greece, Portugal and Spain (2015–16), the political cost of the crisis, which had initially affected the national party systems, seems to have passed to the government level, shaking the process of government formation, changing the identity of incumbents and, ultimately, undermining the stability of the executives.

A decade into the economic disruption that started in 2008 (Frieden & Walter forthcoming; Pontusson & Raess 2012), a vast research repertoire has shown how the Great Recession has affected democratic life in Southern Europe. A number of works have revealed how the structural reforms and austerity policies implemented in response to the crisis (spending cuts, tax increases, welfare trimming and the freezing of public sector employment) together with their main consequences (an upsurge in unemployment, consumer demand compression and social inequalities) have had an impact on the circuit of democratic representation.

The changes in voters’ preferences, the increase in electoral volatility and the decline in electoral turnout; the spread of new and old forms of protest; the decay of mainstream parties and the appearance of powerful anti-establishment challengers; the growth of distrust in political institutions and the dwindling capacity of parties to channel and represent the positions of their voters are among the dimensions affected by the crisis and discussed in the literature (see, among others, Freire & Lisi 2016; Hernández & Kriesi 2016; Magalhães 2014; Moury & De Giorgi 2015; Torcal 2014). In fact, the impact of the crisis on South European democracies has been so great that we cannot but agree with the conclusion that the Great Recession has ended up affecting the quality of democracy at large: ‘worsening of the economy mainly affects the rule of law, electoral accountability, participation, equality and responsiveness’ (Morlino & Quaranta 2016, p. 626).

This volume aims to update our understanding of the South European electoral epidemic through scrutiny of the most recent national parliamentary elections and the most successful challenger parties. It also seeks to highlight the way in which the electoral malaise is being transferred to the government level, with presently unforeseeable consequences for the long-term stability of South European democracy. These goals are reflected in the structure of this introductory article, which pursues a causal progression, examining first the new electoral landscape, then the reshaped party systems to which this has given rise and, finally, the latter’s consequences for government formation and longevity.

What makes this story even more striking is that it is playing out in a region that, in the decades since the fall of the authoritarian regimes in the 1970s, has been widely regarded as a democratic success story and a byword for democratic stability and predictability. It is also worth noting that the 2015–16 elections were held at a time when it seemed the worst of the economic crisis might be over. Portugal had exited its European Union/International Monetary Fund (EU/IMF) sovereign bailout in May 2014, just a few months after Spain exited its bank bailout. In 2014, after several years of recession, all three economies saw a switch to positive growth (of 0.8 per cent of gross domestic product [GDP] in Greece, 0.9 per cent in Portugal and 1.4 per cent in Spain). This was accompanied by a marginal drop in the unemployment rate (by 1.0 per cent of the economically active population in Greece, 1.6 per cent in Spain and 2.3 per cent in Portugal) (European Commission 2015). Yet this muted brightening of the economic prospects did not ameliorate the political discontent generated during the crisis years. Instead, the political crisis continued to deepen.

The advance of the electoral epidemic

Each of the five South European national parliamentary elections that took place in 2015–16 is covered in detail in an article in this volume (Tsirbas 2016 on the Greek contest of January 2015; Tsatsanis & Teperoglou 2016 on Greece in September 2015; De Giorgi & Santana-Pereira 2016 on Portugal in October 2015; Orriols & Cordero 2016 on Spain in December 2016; and Simón 2016 on Spain in June 2016). The aim of this section is not to repeat the details of the individual contests that feature in these articles. Rather, it aspires to offer a comparative overview of the new electoral landscape that has emerged in Greece, Portugal and Spain. To do so, it is necessary also to take into account the four other national parliamentary elections held during the crisis period (Spain and Portugal in 2011, Greece in May and June 2012). Doing so will help us to contextualise the most recent electoral wave in the sequence of crisis elections.

If we look at the electoral panorama of Greece, Spain and Portugal from a late 2016 viewpoint, the first feature to be emphasised is the variation in the number of elections. Five elections took place between 2009 and 2015 in Greece, four between 2008 and 2016 in Spain and three between 2009 and 2015 in Portugal. Since the beginning of the crisis, none of the three countries has avoided early elections. But in Greece and Spain some elections were very close to one another – an indicator of crisis according to Morlino (1998, p. 85) – while Portugal only recorded one snap poll in 2011.

According to the classic formulation by Hirschman (1970), voters have a choice between ‘exit’, ‘voice’ and ‘loyalty’. If we examine the figures on abstention in Table 1, it is clear that in Greece there has been a significant increase in voters opting for ‘exit’. The proportion of those not going to the polls has increased steadily at every Greek election since the start of the crisis. Over the six years between the elections of October 2009 and September 2015, the total increase in abstainers reached 14.7 per cent of the total electorate. Particularly striking is the jump of 7.7 per cent between January and September 2015. If we consider that in this last election 2.4 per cent of those voting chose to cast blank or invalid ballots, this means that less than 54 per cent of voters were prepared to indicate a preference for one of the political parties. If the level of electoral participation is regarded as an important gauge of support for a democratic political system, Greece is currently only narrowly above the 50 per cent mark of systemic legitimacy. While there has also been some rise in ‘exit’ in Spain, the total increase (7.3 per cent) between 2008 and 2016 was half that which occurred in Greece. Moreover, two-thirds of the Spanish electorate, a large majority, continue to go to the polls. In Portugal, there is a perennial problem with the electoral registers, which means that the apparent totals for abstention are not generally regarded as accurate. Nevertheless, what the Portuguese data do suggest is that any increase in abstention during the crisis period has been rather limited.

Table 1. Crisis-elections indicators in Greece, Spain and Portugal (2008–16).

	Greece				Spain			Portugal
	May 2012	June 2012	January 2015	September 2015	2011	2015	2016	2011	2015
Abstention (%)	34.9	37.5	36.1	43.8	31.1	30.3	33.5	42	44.2
	2009 = 29.1				2008 = 26.2			2009 = 40.3
Traditional bipartyism (% of votes)	PASOK+ND	PASOK+ND	PASOK+ND	PASOK+ND	PSOE+PP	PSOE+PP	PSOE+PP	PS+PSD	PS+ PaF
	32.0	41.9	32.5	34.4	73.4	50.7	55.6	66.7	70.7
	2009 = 77.4				2008 = 83.8			2009 = 65.7
TEV	48.5	18.7	20.5	8.4	17.0	35.5	5.5	13.7	13.8
	2009 = 10				2008 = 5.3			2009 = 9.1
ENEP	8.95	5.20	4.43	4.51	3.34	5.83	5.03	3.9	4.6
	2009 = 3.16				2008 = 2.79			2009 = 4.1

Sources: Abstention, traditional bipartyism – Greece: http://www.ypes.gr/el/Elections/NationalElections/Results/; Spain: http://www.infoelectoral.mir.es/infoelectoral/min/; Portugal: www.cne.pt and for 2015 http://www.eleicoes.mai.gov.pt. ENPP– Greece and Spain: Gallagher (2017) (election indices data-set: http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/index.php); Portugal: Freire (2016, p. 185). TEV: Emanuele (2015).

Notes: Figures in bold indicate the most extreme score for each country.

A similar picture, with dramatic or fairly extensive change in Greece and Spain but without significant change in Portugal, emerges when we look at the ‘loyal’ voters: those who not only kept on participating in the long sequence of elections but also kept on voting for the mainstream option, i.e. for the two traditional parties of government. Prior to the crisis, each of these three countries had a well-established two-party system in which, for some decades, two mainstream parties had been the choice of the majority of the electorate. From the figures on traditional bipartyism in Table 1, it is immediately clear that the fall in support for the old governing parties has been overwhelming in Greece. Overall, backing for PASOK (Πανελλήνιο Σοσιαλιστικό Κίνημα – Panhellenic Socialist Movement) and ND (Νέα Δημοκρατία – New Demoracy) has fallen from two-thirds of votes in 2009 to just 34.4 per cent in 2015. The biggest shift occurred at the May 2012 election, which also marked the historic low point in the combined vote share of these two parties (32 per cent).

In Spain, between 2008 and 2016 the combined vote share for the PSOE (Partido Socialista Obrero Español – Spanish Socialist Workers’ Party) and the PP (Partido Popular – Popular Party) slumped from 83.8 to 55.6 per cent, a plunge of almost 30 per cent. The bottom level, however, was hit in the 2015 elections, when traditional Spanish bipartyism attracted less than 51 per cent of votes. In contrast, Portugal has appeared to be going against the trend, since between 2009 and 2011 the traditional main parties slightly increased their vote share. Meanwhile, the picture for 2015 is obscured by the fact that the centre-right PSD (Partido Social Democrata – Social Democratic Party) did not run autonomously but ran together with its government coalition partner, the CDS–PP (Centro Democrático Social–Partido Popular – Democratic and Social Centre–Popular Party),¹ in the PaF (Portugal à Frente – Portugal Ahead) electoral alliance. If the CDS–PP vote share is added to that of the two main parties for 2009 (making a total of 76.1 per cent) and then compared with the combined vote share of the PaF alliance and the PS (Partido Socialista – Socialist Party) in 2015, the result is a drop of about five percentage points. This is clearly not comparable with the extensive realignments in voter choices in Greece and Spain.

Portugal is the ‘exceptional case’ also when we look at the voters who during the crisis years have expressed ‘voice’, changing their voting behaviour and contributing to the entry of new parties into the parliamentary arena. Total electoral volatility (TEV) is an index that measures the minimum percentage of voters who shift their vote between consecutive elections. The data presented in Table 1 show that electoral volatility in Greece peaked to a maximum in the May 2012 election (48.5) – from 10 in 2009 – before slowly declining in the following polls. A spike also characterised the Spanish elections of December 2015, when TEV jumped from 17 to 35.5, the second highest score in the country’s history, before plunging to a historic minimum in June 2016. While in Greece (about) a half and in Spain an (abundant) third of the voters changed their party choices in a specific election, in Portugal the index of volatility remained fairly stable between 2011 and 2015 and only slightly above the mean value (12.2) for the 1976–2009 democratic period.

Finally, the effective number of parties at the electoral level (ENEP) is a measure that helps to gauge the fragmentation of the electoral scene before the translation of votes into seats.² The information provided in Table 1 shows that in Greece the level of fragmentation rose steadily between 2009 and May 2012 (+183 per cent), to decline and somewhat stabilise in the successive elections. This also occurred in Spain, where the biggest increase took place in the December 2015 election (+75 per cent), to decline afterwards. In Portugal, instead, electoral fragmentation remained substantially stable between 2009 and 2015 (+12 per cent) with a minor increase between 2011 and 2015. Overall, while electoral competition in Greece and Spain has become more fragmented, with growing portions of voters supporting a larger number of parties, in Portugal changes have been minimal.

The impact of the electoral epidemic in Greece and Spain is clearly different from the Portuguese case. Both in Greece and in Spain, the lowest level of vote for the traditional mainstream parties, the peak TEV scores and the highest scores in the number of parties at the electoral level all coincide in a single election – that of May 2012 in Greece and of December 2015 in Spain. In 1955 O. Key introduced the concept of ‘critical election’ to signal an election characterised by ‘a realignment within the electorate both sharp and durable’ (p. 11). Key’s concept, based on American elections, did not focus on the decline in vote share for the two main parties or the entrance of new parties into the parliamentary arena. However, we think that it can be helpful to define specific turning points in the electoral sequences that have taken place in Southern Europe.

While we cannot be sure (yet) of the future persistence of the new electoral realignments, the fact that in both Spain and Greece, the three measures all register their highest/lowest scores in the same election allows us to consider these two countries as characterised by the presence of a ‘crisis-critical election’ – a decisive turning point in the sequence of elections that followed the outbreak of the economic crisis.³ In Portugal such a crisis-critical election did not take place. Traditional bipartyism and electoral fragmentation remained more stable than in the other two countries and volatility maintained a level close to the average for the democratic period.

The reshaping of the party systems

Party system transformation is evident from the effective number of parliamentary parties (ENPP), a figure that measures the fragmentation of a party system and the relative size of its parties (see note 2). The data in Table 2 replicate the trends we found for ENEP. Strong rises for Greece in May 2012 (+86 per cent) and for Spain in December 2015 (+74 per cent) were both followed by a decline in subsequent elections. Meanwhile in Portugal fragmentation remained stable between 2009 and 2015, with a slight decline in 2011. Overall, while the Greek and Spanish party systems have become more divided and characterised by smaller political forces, in Portugal nothing substantially changed between 2009 and 2015.

Table 2. Party-system change indicators in Greece, Spain and Portugal (2008–16).

	First three partiesvote share	ENPP	Ideological distance^a between two main parties	Ideological distance between most extreme parties
Greece
2009	PASOK–ND–KKE	2.59	2.3	5.9 (LAOS–KKE)
2012	ND–SYRIZA–PASOK	4.83–3.76^b	4.3	7.5 (XA–KKE)
2015	SYRIZA–ND–XA	3.09–3.24^c	4.5	7.5 (XA–KKE)
Spain
2008	PSOE–PP–IU	2.36	3.8	5.4 (PP–IU)
2011	PP–PSOE–IU	2.60	3.8	5.4 (PP–IU)
2015	PP–PSOE–PODEMOS	4.53	3.9	6.0 (PP–Podemos)
2016	PP–PSOE–UP	4.16	3.8	6.2 (PP–UP)
Portugal
2009	PS–PSD–(CDS–PP)	3.3	2.5§	5.1 (CDS–CDU)^d
2011	PSD–PS–(CDS–PP)	2.9	2.9	5.6 (CDS–BE)
2015	(PSD–CDS)–PS–BE	3.3	3.7	6.0 (PaF–CDU)

Notes:

Sources: Spain: Cis (2008); Simón (2016); Portugal: Freire (forthcoming); Greece: Hellenic Voter Study 2009, Hellenic Voter Study 2012 (www.elnes.gr) and Hellenic National Election Voter Study 2015 (http://doi.org/10.3886/E100074V16); see also Andreadis et al. (2015).

^a Ideological distance refers to political parties’ left–right position according to the electorate’s perceptions on a left–right scale that ranges from 1 (extreme left) to (10 extreme right).

^b Data refer, respectively, to the May and June elections.

^c Data refer, respectively, to the January and September elections.

^d Data refer to 2008.

Another indicator of change is whether the first three parties in each system remain the same. As shown in Table 2, change occurred in all three countries. In Greece, only one of the first three parties in 2009 remained in the top three six years later. The Greek Communist Party (Κομμουνιστικό Κόμμα Ελλαδάς – KKE) was relegated to a lower ranking in 2012 followed by PASOK in 2015. They were replaced by the radical-left SYRIZA (Συνασπισμός Ριζοσπαστικής Αριστεράς – Coalition of the Radical Left) and the neo-nazi GD (Χρυσή Αυγή – Golden Dawn). In Spain, the traditional party of the radical left, IU (Izquierda Unida – United Left), was ousted by the newcomer Podemos (We Can) in 2015, and in 2016 the two joined forces in an electoral alliance. This union, however, did not work as expected and Unidos Podemos (Together We Can – UP) failed to seize the second place. In Portugal, the 2015 election saw the radical-left BE (Bloco de Esquerda – Left Bloc) emerge as third party.

In all three cases, the management of the economic crisis and the emergence of new challengers increased polarisation within the party system. In the eyes of Greek voters, the rise of SYRIZA (which replaced PASOK as second party in 2012 and became first party in 2015) doubled the ideological distance between the two main parties. As can be seen from Table 2, in 2009 the electorate saw PASOK and ND as being relatively close together, with only half the post-2012 ideological distance between ND and SYRIZA. Meanwhile, the parliamentary breakthrough of the neo-nazi GD increased the perceived distance between the parties at the two ends of the left–right spectrum. In both Spain and Portugal, the two main parties remained the same. In Spain, voters saw the distance between the PP and the PSOE remain the same as both parties moved to the right (Simón 2016). In Portugal the gap between the PSD and the PS was seen as widening in 2008–15 due to a pronounced shift to the right by the PSD during its government term (Freire forthcoming; Lisi 2016). Both Spanish and Portuguese voters also perceived an increased distance between the most extreme parties of the political spectrum, the extremes being seen as stretching the ideological space.

Table 2 also casts further light on the decline of traditional bipartyism. In both Spain and Greece, the crisis period has squeezed support for both of the traditional governing parties. However, the mainstream right has displayed significantly greater resilience than the socialist left. The Greek PASOK, shrunk to a fraction of its former strength, was relegated to minor party status and definitively ousted from its position as the main political force on the left. In Spain the PSOE has so far remained second party but suffered a snowball decline that halved its electoral support between 2008 and 2016. The fate of these once powerful South European socialist parties is consistent with the general picture of a crisis of social democracy in Europe (e.g. Keating & McCrone 2013). Only in Portugal does the socialist party stand out as a case of resilience. In contrast to PASOK – entangled in government from 2009 to early 2015 – the 2011 electoral defeat that ousted the PS from power also shielded the party from austerity-governing exhaustion. Unlike the PSOE – entangled in internal party rifts – the Portuguese socialist organisation was more cohesive and able to renew and rejuvenate its leadership twice after 2009. These key preconditions proved very important in 2015 when the centre-right coalition lost the majority in parliament.

Meanwhile, the parties of the traditional communist left have maintained their party bases but have not benefited from the crisis. Indeed, the KKE has lost electoral support while its Iberian counterparts have made, at best, marginal gains. In all three cases, these parties’ vote shares remain in single figures. Despite their electoral marginality, however, both the PCP (Partido Comunista Português – Portuguese Communist Party)⁴ and IU came to play an important role after the 2015 elections. The fact the KKE did not was due to its own strategic choice to remain aloof from any government coalition.

The decline of old parties has opened up space for new contenders. The challengers were not confined to a specific section of the political spectrum. Particularly striking and common to all three countries was the rise of the non-communist radical left, turning SYRIZA, Podemos and the BE into key players in government formation in 2015–16. The prominence of the radical left differentiated these three South European countries from broader trends in Europe, where it was mainly the radical right that benefited from popular dissatisfaction. However, the far right was very visible in Greece, where two new contenders made parliamentary breakthroughs in 2012. Both parties acquired more significant roles in 2015, when the neo-nazi GD emerged as the third largest political force while the nationalist ANEL (Ανεξάρτητοι Έλληνες – Independent Greeks) became a junior government partner. But apart from the rise of new forces at the two ends of the political spectrum, there also appeared a new challenger centre. In addition to the less successful To Potami (the River) and Enosi Kentroon (Union of Centrists) in Greece, its most prominent incarnation is Ciudadanos (Citizens) in Spain, which in 2015 achieved a breakthrough from regional party to significant national parliamentary player. In this volume, we have chosen to focus on five major challenger parties (Tsakatika 2016 on SYRIZA; Rodríguez Teruel & Barrio 2016 on Ciudadanos; Rodríguez Teruel, Barrio & Barberà 2016 on Podemos; and Lisi 2016 on the BE and the PCP) that became key players in the processes of government formation following the 2015–16 elections.

Governing before the crisis

In the period following the fall of the three dictatorships and before the outbreak of the eurozone sovereign debt crisis in late October 2009, all three countries – with one brief period of exception in Greece and a longer one in the Portuguese case – followed a predictable process of government formation leading to an expected outcome.⁵

In Greece, the first democratic elections following the fall of the dictatorship were held in November 1974. From then until the outbreak of the eurozone sovereign debt crisis, a total of 13 parliamentary elections were held. On no occasion was there a need for repeat polls, as every election resulted in the formation of a government with a parliamentary majority. Only three of these governments completed a full four-year term. In almost all the cases where early elections were called, this was due to the choice of the prime minister of the day to hold early elections, choosing a time they deemed more propitious for their own party. The only exception was the ND government of 1990–93, which was brought down by the defection of a group of its members of parliament (MPs), triggering new elections. Although there were cabinet reshuffles during the lifetime of most Greek governments, there was never a case in which a parliamentary majority broke down, resulting in the formation of a new government supported by a different constellation of parties from the same parliament. As Table 3 shows, 11 of these 13 elected administrations were single-party majority governments, with the two main parties, PASOK and ND, alternating in power. So the identity of the prime minister was always known from election night and no interparty negotiations were required to form a government. Thus, Greece in the period from the fall of the dictatorship to the fall of the crisis generally presented a picture of stability and predictability. The only issue was which of the two main parties would emerge first in the election.

Table 3. Government composition in Greece, Spain and Portugal (1974–2016).

Greece	Spain	Portugal
1974 Single-party majority ND	1977 Single-party minority UCD	1976 Single-party minority PS
1977 Single-party majority ND	1979 Single-party minority UCD	1978 Majority coalition PS–CDS
1981 Single-party majority PASOK	1981 Single-party minority UCD	1978 Non-partisan*
1985 Single-party majority PASOK	1982 Single-party majority PSOE	1978 Non-partisan*
1989a Majority coalition ND–SYN	1986 Single-party majority PSOE	1979 Non-partisan*
1989b Majority coalition ND–SYN–PASOK	1989 Single-party majority PSOE	1979 Majority coalition PSD–CDS–PPM
1990 Single-party majority ND	1993 Single-party minority PSOE	1980 Majority coalition PSD–CDS–PPM
1993 Single-party majority PASOK	1996 Single-party minority PP	1981 Majority coalition PSD–CDS–PPM
1996 Single-party majority PASOK	2000 Single-party majority PP	1983 Majority coalition PS–PSD
2000 Single-party majority PASOK	2004 Single-party minority PSOE	1985 Single-party minority PSD
2004 Single-party majority ND	2008 Single-party minority PSOE	1987 Single-party minority PSD
2007 Single-party majority ND	2011 Single-party majority PP	1991 Single-party majority PSD
2009 Single-party majority PASOK	2015 No government formed	1995 Single-party minority PS
2011 Majority coalition PASOK–ND–LAOS	2016 Single-party minority PP	1999 Single-party minority PS
2012a Νο government formed		2002 Majority coalition PSD–CDS
2012b Majority coalition ND (with external support of PASOK–DIMAR)		2004 Majority coalition PSD–CDS
2013 Majority coalition ND–PASOK		2005 Single-party majority PS
2015a Majority coalition SYRIZA–ANEL		2009 Single-party minority PS
2015b Majority coalition SYRIZA–ANEL		2011 Majority coalition PSD–(CDS–PP)
		2015 Single-party minority PS (with external support of BE–PCP)

Notes:

Date in bold indicates that the formation of the government took place after elections.

* Between August 1978 and December 1979, General Ramalho Eanes, the first President of the Portuguese Republic, aiming to enlarge his institutional role, appointed three prime ministers who represented not a political party but the President’s own initiative. The first of these governments, led by Nobre da Costa, was rejected by parliament; the second, led by Mota Pinto, lasted for a few months but could not pass the budget and had to resign; finally, the third, led by Lourdes Pintasilgo, was installed to preside over new elections (Lobo 2005, pp. 82–83).

The exception to this rule were the two elections of 1989, when a change in the electoral law made single-party government formation harder. The result were two elections in short order (June and November), each followed by the formation of a short-term coalition government to achieve a specific mission (in the first case, to clean up after political scandals and, in the second, to tackle urgent economic problems). The first coalition stood out for bringing together the two significant party actors with the largest ideological distance between them, the right-wing ND and the radical left SYN (Συνασπισμός της Αριστεράς και της Προόδου – Coalition of the Left and Progress). The second was unusual as a so-called ‘ecumenical government’ including all three major parliamentary forces (ND, PASOK and SYN) and leaving only three MPs in opposition.⁶ After this nine-month interlude of extraordinary coalitions, Greece returned to the usual pattern of single-party majority government, which then prevailed for the next 18 years.

Like Greece, Spanish democracy in the pre-crisis period also showed remarkable political stability, with no need to repeat polls and a smooth succession of single-party governments. In the period between the first democratic election in June 1977 and the outbreak of the eurozone sovereign debt crisis in October 2009, only ten elections were held and 11 governments were installed. Each government lasted until a new parliament was elected, with the only exception of two UCD (Unión de Centro Democrático – Union of the Democratic Centre) cabinets formed in the 1979–82 parliament. Unlike in Greece, where all governments could rely on a majority of seats, in Spain only four of the 11 cabinets could count on such a strong base of parliamentary support.⁷ Spanish prime ministers, however, have been able to form and lead minority single-party governments thanks to the external support of several regional parties. In contrast to Greece and Portugal, Spain is the only country that has never experienced any coalition government and (so far) has a 100 per cent record of single-party governments. After the early 1980s collapse of the UCD, the party that had led the transition to democracy and ruled the country until 1982, the same political forces, PSOE and PP, alternated in power without interruption.

In Portugal, in contrast, alternation in power between the two main parties, the PSD and the PS, is a more recent achievement, since it was only after 1987 that single-party executives were formed according to a predictable pattern. As Table 3 shows, the April 1976 elections–the first held after the approval of the Constitution – opened up a period of high political instability. Between 1976 and 1987, in fact, five parliamentary elections resulted in the formation of ten governments, supported by different parties. As the orthodox-communist PCP was not considered a potential governing partner by the socialists (see below), the PS preferred to set up a minority cabinet (1976–77) and form coalitions with the right-wing CDS (1978) or the centre-right PSD (1983–85). The PSD, in turn, apart from governing with the PS in the mid-1980s, also formed a coalition with the tiny monarchist party (PPM) and the CDS (1979–83) as well as a minority single-party government (1985–87). The picture is then completed by three short-lived presidential cabinets that failed to bring any stability.

It was only with the electoral realignment of 1987 that the PSD and PS became the government-makers of the Portuguese system. Accordingly, the seven elections held between 1987 and 2009 resulted in only eight governments, five of which rested on an absolute majority of seats. There was only one legislature with two governments and this was due to the necessity to choose a new prime minister after José Manuel Durão Barroso resigned to preside over the European Commission in 2004. In the same legislature the PSD resumed the habit of coalescing with the CDS–PP to control the majority of seats.

In short, the Portuguese government formation process before the crisis showed two radically different phases. A first one characterised by ‘trial and error’ (coinciding with the gradual process of democratic consolidation experienced by the country) and a second one marked by a predictable and stable pattern of alternation between governments led by the two main parties. The small programmatic and ideological distance between the PSD and CDS–PP made an alliance on the centre-right readily available, while the PS did not have such an opportunity because of the huge political distance from the communists. The PCP has long been a Marxist–Leninist force that has defended the heritage of the 1974 Carnation Revolution, rejecting the consolidation of a liberal-democratic regime in Portugal as well as the country’s entry into the European Community. The anti-system identity of the communists, in turn, led to the setting up of an ‘inner party system’, that is a party system in which the other political forces collaborated and redefined the terms of reciprocal competition so as to marginalise the PCP (Bosco 2000, 2001). This explains why a government coalition including the PS and the PCP was not formed despite the absolute majority of seats gained by the two parties in the 1976, 1983 and 1995 elections. Throughout the pre-crisis period, the communists’ participation in a national government was simply not an option for the Portuguese socialists.

The outbreak of the government epidemic

Following the outbreak of the eurozone sovereign debt crisis in late 2009, the economic shock gradually spread to the political systems of the three South European countries, with a speed and strength determined in each case by the extent of the national economic problems, the severity of the measures adopted to address them, the timing of elections, and the characteristics of the government and opposition parties. In Greece, Spain and Portugal the initial phase of the crisis had an immediate impact at the electoral and party-system level – the first phase of the ‘electoral epidemic’ discussed above. Several of the ‘electoral epidemic’ features clearly marked the Spanish, Portuguese and Greek elections of 2011–12.

However, in the Greek case this early crisis period also had an impact at the government level. As the country worst affected by the crisis, Greece had already experienced two EU/IMF bailouts by spring 2012. It was the only one of our case study countries to experience significant government formation problems during this period. It also underwent what currently appears to be a permanent switch to a system of coalition government, overturning the dominant pre-crisis model of single-party majority governments. At this point, Greece looked like an exceptional case. In comparison, Spain and Portugal were less obviously affected and showed more traditional patterns of government formation.

Greece 2012: the government epidemic begins

In Greece, the first manifestation that government formation was entering a new phase occurred in November 2011, less than a year and a half after the country’s first EU/IMF bailout in May 2010. This was the first occasion in post-dictatorship Greece when an elected government essentially collapsed. The demise of the Papandreou government was due to the dramatic loss of legitimacy generated by the implementation of tough austerity measures. It was triggered by the intervention of EU leaders to avert a referendum in Greece on the planned second bailout. This was followed by a process entailing a series of ‘firsts’ (see Table 4). The first unprecedented act was a parliamentary vote of confidence given, not for the government to continue, but as a mandate to negotiate its replacement (Verney & Bosco 2013, p. 405). Several days of interparty negotiations concluded with the formation of the first coalition government since 1990. It was also the first occasion in post-dictatorship Greece that the same parliament had produced a second government based on a different constellation of party support.

Table 4. The government epidemic in Greece (2011–13).

Election	Unprecedented events during government formation	Uncharted government	Repeat
		outcomes	elections
November 2011	Collapse of elected government
	Parliamentary vote of confidence as mandate to dissolve government and form a new one
	Process of direct talks among party leaders chaired by President of the Republic (instead of three-day mandates to leaders of three major parties)
		Formation of new government from same parliament	No
		Inclusion of radical-right government partner
May 2012	First party handed back mandate within 24 hours
	‘Fourth round’ included direct consultation of President of the Republic with neo-nazi leader	Failure to form a government	Yes after six weeks
June 2012	None	Agreement to form a long-term coalition government	No
		Two of coalition partners did not participate directly in government
June 2013	None	First coalition between the two traditional parties of government without presence of a third party	No

The outcome was the formation of a three-party coalition of PASOK, ND and the radical-right LAOS (Λαίκός Ορθόδοξος Συναγερμός – Popular Orthodox Rally) . The only previous occasion on which there had been an oversize coalition was the short-term government formed in November 1989. But this time there was a significant difference: it was the first time since the fall of the Junta that the far right had participated in a Greek government. Meanwhile, although this was a government of party politicians, it was led by a technocrat, former governor of the Bank of Greece and vice-president of the European Central Bank, Lucas Papademos. This had only occurred once before, in the case of the November 1989 government, which was also headed by a former Bank of Greece governor. As in 1989–90 the coalition agreement was for a short-term government to handle a specific mission, in this case the negotiation of the second Greek bailout and the partial restructuring of the national debt. The ratification of the new bailout agreement shook the government to the core. LAOS withdrew before the parliamentary vote in February 2012, in which a total of 43 PASOK and ND MPs voted against the bailout while a further 22 did not appear in parliament. Nevertheless, the government retained a parliamentary majority and remained in power until the debt haircut was complete. In the end, the government lasted just under six months.

The ‘earthquake elections’ (Teperoglou & Tsatsanis 2014) of May 2012 produced a shocking fragmentation of the vote and broke the mould of the two-party system. The outcome was a seven-party parliament, something not experienced since 1977, when the party system was in a process of reconstruction after the dictatorship. But, unlike in 1977, this parliament had no major parties. The largest, ND, won less than 19 per cent of the vote. Even with the 50-seat bonus given to the largest party, it held only 108 out of 300 seats. But the way in which the seats had been distributed among the parties was not the only problem.

Coalition options were reduced by the presence of two anti-system parties: the KKE, which rejected any suggestion of government participation, and the neo-nazi GD, which all other parties rejected as a potential government partner. Ideological distance was therefore an even more significant limitation. With the country on the verge of ‘Grexit’, the basic question was whether Greece should implement the second bailout in order to stay in the eurozone. The gap between parties that disagreed on this fundamental question was simply unbridgeable. Another major difficulty was the sweeping popular rejection of the political elites indicated by the election. Almost one in five votes (19 per cent) had been cast for parties that failed to reach the three per cent parliamentary threshold and were therefore wasted. This meant an unusually extensive disconnect between the parliamentary arithmetic and the electoral arithmetic. The only politically feasible coalition, between ND, PASOK and DIMAR (Δημοκρατική Αριστερά – Democratic Left), which all prioritised keeping Greece in the eurozone, would have enjoyed an 18-seat parliamentary majority. But with only 38.2 per cent of the vote, it would have been very far from a popular majority and hence would have lacked legitimacy.

Compounding the difficulties was the strict three-day time limit placed on the mandates for government formation offered in turn to each of the three largest parties. In fact, ND returned its mandate in less than one day and the whole process concluded nine days after the election. Because of the presence of anti-system parties, during the fourth round of government formation talks the President of the Republic initially consulted with individual party leaders before calling a meeting with those he felt might be able to cooperate. The other party leaders thus avoided meeting with the head of GD, who did not take part in direct talks on government formation. Nevertheless, the sight of the neo-nazi leader being received at the presidential palace was a shock. Given the parliamentary seat distribution, it is hardly surprising that the outcome was an unprecedented failure to form a government, followed by an equally unprecedented recourse to repeat elections six weeks later.

In contrast, following the June 2012 election, the government formation process was smooth and fast. The election returned the same seven parties to parliament but the big drop in wasted votes meant these parties now represented 94 per cent of those who had voted compared with 81 per cent in May. The government formed, a three-party coalition of ND, PASOK and DIMAR and led by ND leader Antonis Samaras, held a parliamentary majority of 27 seats and represented 48 per cent of the popular vote. The coalition agreement was announced three days after the election, before the expiry of the first mandate, which had been given to the lead party, ND. While this government too had a special mission – to keep Greece in the eurozone – it was the first Greek coalition to be formed, not as a short-term expedient with a predetermined endpoint, but with a longer-term perspective.

This was the first (and so far the only) Greek coalition in which two of the supporting parties chose not to take up ministerial posts. DIMAR nominated two independent personalities to ministerial positions and two technocrats as deputy ministers, but did not participate with party cadres, while PASOK did not take direct part in the government at all. Instead, these two parties supported the government from outside after offering an initial vote of confidence. While the government was a mix of politicians and technocrats, it was essentially dominated by MPs from ND.

The coalition broke up after one year following ND’s decision to close down the public broadcaster within 24 hours, despite the opposition of its coalition partners. Three days later a new coalition agreement was announced between ND and PASOK, again led by Samaras. This was the second time during the crisis that a new government, supported by a different constellation of parties, had been formed from the same legislature. It was also the first time that the two traditional governing parties had formed a coalition without the mediating presence of a third party. Although an independent technocrat was retained as finance minister, the rest of the ministerial posts were shared out among politicians from the two governing parties. This second government to emerge from the legislature formed in June 2012 lasted for almost two-and-a-half years and can be regarded as normalising the experience of coalition government in Greece. Nevertheless, despite offering relative stability compared with what had preceded it, the government lasted only just over half the normal length of a parliamentary term.

It is worth considering that 2007–15 would have been the normal life-span of just two Greek governments, serving four years each. Instead, this period saw six governments: two single-party governments of ND, one single-party PASOK government and three different coalitions. A final point worth making is that, despite the difficulties of government formation and the move to coalitions, there was continuity with the pre-crisis period in the sense that the two traditional parties of government continued to exercise government power.

Spain and Portugal 2011: business as usual

In Spain the process of government formation that followed the November 2011 elections was velvet smooth and carried out without departing from the traditional pattern. As the PP had won the second largest majority ever (186 seats out of 350, following the 202 seats gained by the PSOE in 1982), a strong single-party government was installed under the leadership of Mariano Rajoy. Rajoy remained in charge until the very end of the legislature (in 2015) and was one of the few Spanish prime ministers not to resort to an early dissolution of parliament.⁸ This was a marked contrast with the rough departure of the previous PSOE government. Re-elected as prime minister in 2009 after a successful first term, José Luis Rodríguez Zapatero in his second term was forced to shift from fiscal expansion to fiscal retrenchment. The change proved devastating and, in a climate of political protest and plunging popularity, Zapatero declined to run for a third term and called elections for November 2011, four months early.

In Portugal, as in Greece, the elections took place after a bailout. The Memorandum of Understanding on financial assistance to the country was in fact signed by the three main parties (PS, PSD and CDS–PP) in May 2011, less than a month before the elections took place on 5 June. The outgoing socialist prime minister, José Sócrates, lost at the polls and the PS moved into opposition while the PSD emerged as the largest party, but without a majority of seats. The PSD leader, Pedro Passos Coelho, therefore negotiated a coalition pact with the CDS–PP, in order to install a strong and stable government that could face the challenges linked to the forthcoming economic adjustments. Such a coalition was not new, as the parties had governed together more than once (see Table 3).

In short, the installation of the 2011 government did not depart from the beaten track. According to Magalhães (2012) the only novelties concerned three features that could be traced to the urgent measures agreed with the creditors. The first was the unusually short time needed to take office – less than three weeks after the polls. The second was the small size of the new government, with just 12 ministers and 35 junior ministers. Finally, the third feature was the significant presence of independents in the cabinet. This latter feature was ‘symptomatic … of the perceived need to insulate governments from pressures from party organisations, and of the level of hostility towards parties which prime ministers intuit in public opinion’ (Magalhães 2012, p. 324). Eventually, such a design helped the PSD–CDS coalition to last until the end of its mandate, in 2015 – so much so that it has been considered ‘the most stable coalition in the history of Portuguese democracy’ (Freire 2016, 184). Compared with Greek cabinets, the centre-right governments installed in 2011 in Spain and Portugal thus shone in their durability.

Charting new territories

As we have seen, in 2011 Spain and Portugal had experienced the spread of the electoral epidemic but without any unusual government consequences. In contrast, unprecedented events during (multiple) processes of government formation, uncharted outcomes in the identity of the governing parties and the necessity to repeat elections were three features that defined a new syndrome that made its first appearance in Greece. With the intensification of the electoral epidemic in the elections of 2015–16, the ‘government epidemic’ now began to spread to Spain and Portugal as well. Moreover, with the new election season in Greece, Spain and Portugal, the radical left emerged as a major actor in government formation in all three countries.

Greece 2015: standard process, uncharted outcome, repeated elections

In Greece, it had been widely anticipated that the election of January 2015 would be followed by a difficult process of coalition negotiation. In particular, on election night there was widespread concern that the rotation of three-day mandates among the three major parties would result in a mandate to form a government being given to the neo-nazi GD, which had emerged as third party. This would have been a very negative development for Greek democracy. However, things did not happen that way. SYRIZA won 149 seats, just two seats short of a parliamentary majority. It was clear that the only possible outcome of this election could be a coalition government led, for the first time in Greek history, by a radical-left party. (While SYRIZA’s forerunner, SYN, had participated in the coalitions of 1989–90, this had been as a junior partner.) It was also to be the first time since 1974 that a Greek government had been formed that did not include either PASOK or ND (see Table 5). The January 2015 government formation process was uneventful. The morning after the election, after a one-hour meeting between the two party leaders, SYRIZA and ANEL announced their agreement to form a government. The coalition represented 41.1 per cent of the popular vote and enjoyed an apparently safe parliamentary majority of 12 seats.

Table 5. The government epidemic in Greece (2015).

Election	Unprecedented events during government formation	Uncharted government outcomes	Repeat
			elections
January 2015	Speed of the coalition agreement (announced day after the election)	First Greek government led by a radical-left party
		First government since 1974 not to include either of the two traditional main parties
		Coalition between radical left and nationalist right	Yes after eight months
September 2015	No	No (previous coalition repeated)	No

For the previous few years, coalition government had become ‘the new normal’ in Greek politics. What raised eyebrows in this case was not the fact of a coalition but its ideological heterogeneity. The nationalist ANEL was arguably less far to the right than the anti-immigrant and anti-euro LAOS which had participated in the 2011–12 coalition. But it lay towards the opposite end of the political spectrum from the radical-left SYRIZA. What brought the two parties together – and indeed made them logical allies in the context of crisis-era Greek politics – was their agreement around the central issue of Greece’s relationship to the eurozone. Both parties agreed that Greece should remain in the euro but withdraw from the bailout. This was a distinctly different position from those parties which wanted Greece out of both the eurozone and the bailout (KKE, GD) and those which wanted Greece to remain in the single currency and in consequence accepted the bailout (ND, PASOK and the new centre party, To Potami). As a result, however, this apparently solid coalition government was based on a programmatic illusion which was to undermine its stability in a matter of months.

The new government attempted to negotiate a new deal for Greece, entailing a bailout exit accompanied by debt restructuring and a renegotiation of the high budget surpluses that the country was required to produce. This rapidly led to impasse with the creditors. Less than a month after the election, under pressure from a European Central Bank decision limiting Greek financial liquidity, the government signed a commitment to complete the bailout, honour the country’s debts and not to take unilateral measures that had a fiscal impact without consultation. This reversal of SYRIZA’s pre-election promises evoked such a strong reaction from within the party’s parliamentary group that the government did not bring the agreement to parliament for ratification. On 26 June, six months after the election, Prime Minister Alexis Tsipras suspended the negotiations on the last bailout evaluation and the following day called a referendum on the creditors’ proposals. The days before the 5 July vote were marked by a disorderly exit from the second bailout and by Greece becoming the first developed country to default to the IMF.

The 61.3 per cent referendum vote in favour of rejecting the EU/IMF terms, in line with the government’s recommendation, was interpreted by Tsipras as a mandate to negotiate a new three-year bailout agreement. The third bailout, like its predecessor, split the governing coalition – in this case so extensively that the bailout deal only passed through parliament with the support of three opposition parties. The August vote on the bailout terms reduced the coalition to 118 MPs, i.e. 32 short of a parliamentary majority. In contrast to the second bailout, there was no defection of a governing party: ANEL remained loyal to its larger partner. Instead, the split came from within SYRIZA, and the day after the vote 25 SYRIZA MPs officially broke away to form a new party, LAE (Λαική Ενότητα – Popular Unity).

The result was the calling of repeat elections, held on 20 September, just eight months after the previous contest. As in January, pre-poll expectations of a potentially complex government formation process were confounded. Despite the upheaval of the previous months, the two government partners lost less than two per cent of the vote, winning a renewed majority, albeit reduced to five parliamentary seats. The renewal of the coalition agreement was essentially announced just hours after the polls closed, with the joint appearance by the two party leaders, arm in arm on an election platform. This time, just as in the pre-crisis era, Greek voters already knew who their new prime minister would be from election night.

Iberia: the government epidemic spreads

In both Spain and Portugal the outgoing centre-right parties won the 2015 elections, as they gained the largest shares of votes and seats. But they paid a high electoral price for their governing record, both the Spanish PP and the Portuguese PSD–CDS losing the majorities they had enjoyed in parliament. Meanwhile, the PSOE and PS, which in both cases came second, were unable simply to replace the incumbents. In Spain, neither the seats gained by the centre-right and centre parties (PP and Ciudadanos), on the one hand, nor by those on the centre-left and left of the political spectrum (PSOE, Podemos and IU), on the other, were sufficient to form a majority coalition. Two additional – and bigger – obstacles on the road to government formation were the programmatic distance between (especially) the PSOE and Podemos, on the one hand, and the PSOE and PP, on the other, and the outbreak of internal party conflicts (within both the PSOE and Podemos) on the strategies to follow in the new competitive arena. In contrast, in Portugal, the three left-wing parties – PS, BE and PCP – had won an absolute majority of parliamentary seats, a key difference from Spain. Such a PS–BE–PCP majority had occurred only once before, in 2009, when the eurozone crisis had just begun. At that time, the traditional pattern of installing a minority cabinet led by the first party (PS) had been chosen without much discussion.

Overall, in the past, situations of this kind would have led to the installation of minority governments in both countries. This time, however, the outcomes were completely new – and divergent.

Portugal 2015: breaking the mould in process and outcome

In Portugal, the process leading to the installation of the new prime minister after the 4 October elections produced several unprecedented events and a record. It might be noted though that many of the unprecedented events occurred as a result of the Portuguese institutional constraint that prohibits calling fresh elections in the first six months of a new legislature. In the post-bailout era, a caretaker government until April 2016 would have been problematic for Portuguese financial credibility.

The first breach with the past occurred with the election of a member of the second party to the post of president of the Assembleia da República (see Table 6). Considered as the second highest figure in the Portuguese state, the parliamentary speaker had always come from the first party. In 2015, however, the PS, BE and PCP disregarded the consensual rule and stepped in with the weight of their majority in support of a member of the second party. The second pattern-breaking event occurred when, on 30 October, a minority PSD–CDS government was sworn into office under the leadership of the previous prime minister, Passos Coelho. As shown in Table 3, Portuguese coalition cabinets had always enjoyed parliamentary majorities and this was the first time ever that a coalition government lacked such a resource.⁹ The third act in the new path to incumbency occurred on 10 November, when Passos Coelho’s newborn government had to resign after his programme was voted down in parliament by a motion of no confidence presented by the PS and supported by the whole opposition. The rejection was not unexpected, as the PS, BE and PCP had all declared that they would vote against the cabinet. However, it was the first time since 1976 that a party government was rejected in parliament during the discussion of its programme.¹⁰

Table 6. The government epidemic in Portugal (2015).

Election	Unprecedented events during government formation	Uncharted government	Repeat
		outcomes	elections
October 2015	Election of the president of parliament from the ranks of the second party
	Minority coalition cabinet sworn into office
	Government rejected in the parliamentary investiture session	Appointed prime minister steps down after his programme is voted down in the investiture session
	President appoints another prime minister – mandate to the second party
	President requests formal document from opposition leader concerning policies of a possible government	Minority left government backed by radical parties. End of the Portuguese inner party system	No

After the cabinet programme was voted down in the investiture session, the process of government formation had to start again. This time, presidential choices seemed limited to two options: (1) to nominate a PSD–CDS caretaker government that would run the country until new elections were called; or (2) to appoint the leader of the socialist party, António Costa, as prime minister of a minority cabinet backed by the BE and the PCP. President Cavaco Silva, who would have preferred an oversize coalition of PSD, PS and CDS, had declared that a left-wing executive could jeopardise the trust and external credibility Portugal had gained during the previous parliamentary term (Fernandes & Jalali 2017).

Therefore, the fourth unexpected turn in the government formation process was the public request made by Cavaco Silva to the PS leader on 23 November to provide a formal document with clarifications on the agreements he had signed with the BE and PCP. As Costa had not yet been appointed prime minister and the requests – aimed at forging ‘a stable, durable and credible government solution’ – dealt with issues such as the adoption of the budget laws, compliance with international commitments and the stability of the financial system, the presidential initiative was clearly off the beaten track and designed to influence the governmental process.¹¹ Finally, on 24 November – after having received a formal letter from the PS leader with the required clarifications – Cavaco Silva appointed Costa as the new prime minister.

If a mandate to the second party was something completely unprecedented, it paled when compared with the outcome of the process. On 3 December, Costa’s minority government, backed by the BE and PCP, became the first executive since 1976 to receive the support of the radical left. Such a pact – reinforced by the subsequent weekly meetings held among the three parties (Lisi 2016) – amounted to a momentous turning point for the Portuguese political system. It entailed the end of the inner party system formed in 1976 and the legitimation of the PCP as a governing force, a completely uncharted outcome. With a duration of two months after the election, the government formation process was thus not only the longest in Portuguese democratic history (Fernandes 2016, p. 897), but also the most innovative.

Spain 2015–16: disrupted patterns, repeated elections, almost-standard outcome

In Spain the December 2015 electoral results and the ensuing balance of power in the Congress resulted in an unpredictable path to incumbency punctuated by several unprecedented events (see table 7). As in Portugal, the first novelty was the election of a figure from the second party (PSOE) as speaker of parliament. This resulted from a pact between the PP, PSOE and Ciudadanos and broke the rule that a member of the first party would always hold the office. However, the round of consultations made by the King to appoint the candidate prime minister led to a bigger surprise when Mariano Rajoy, the leader of the first party, rejected the nomination on 22 January, adducing that he had not enough support in parliament to win a confidence vote. This was something that had never occurred before. Rajoy’s decision led the King to promote a second round of consultations with the parties and, for the first time since the return to democracy, to appoint the leader of the second party (Pedro Sánchez of the PSOE) as candidate prime minister on 2 February 2016. This was a completely unprecedented event in post-1977 Spain, as the King had never given a mandate to form a government to the second party.

Table 7. The government epidemic in Spain (2015–16).

Election	Unprecedented events during government formation	Uncharted government	Repeat
		outcomes	elections
December 2015	Election of the president of parliament from the ranks of the second party	Outgoing cabinet remains as caretaker government for ten months
	Leader of first party rejects appointment as candidate prime minister
	The King appoints another candidate prime minister – mandate to the second party
	Candidate prime minister rejected in the second vote of a parliamentary investiture session
	Two-month electoral countdown activated
	Multiple rounds of consultations by the King take place
		Failure to form government	Yes after six months
June 2016	The King appoints the leader of the first party as candidate prime minister twice
	Same candidate prime minister submits to two different sessions of investiture
	Candidate prime minister wins investiture vote after having been rejected three previous times
	Main opposition party abstains in investiture vote of prime minister instead of voting against	Candidate prime minister gains investiture vote thanks to the abstention of the main opposition party	No

The traditional pattern of government formation was shaken even further when Sánchez’s investiture was rejected twice by the Congress in the votes held on 1 and 4 March. Under the Spanish Constitution, the candidate prime minister appointed by the King has to pass an investiture vote in the Congress of Deputies. Two votes are allowed. In the first, the candidate needs the support of an absolute majority of MPs (176 out of 350). If this majority is not attained, a new vote takes place within 48 hours, when the candidate only needs the confidence vote of a simple majority of MPs, i.e. those voting yes must be more than those voting no (with those abstaining reducing the quorum necessary for the election). Sánchez’ rejection was a real watershed, as no other candidate prime minister had been defeated in a second investiture vote.

This exceptional event led to another unheard-of development, the start of the ‘electoral countdown’. According to the Spanish Constitution (art. 99.5), if within two months after the first vote of investiture no candidate prime minister obtains the confidence of the Congress, the King must dissolve parliament and call new elections. Thus, after Sánchez’s defeat, the two-month countdown to calling a fresh election – due to end on 3 May – started to influence the negotiations between parties for the first time in Spanish democratic history. As new polls were possible, attention also turned to the opinion surveys and each party’s electoral prospects.

Despite a number of meetings between the PSOE, Podemos, Ciudadanos and other parties, the negotiations failed and it was not possible to set up a base of support for a cabinet led by the PSOE. Acknowledging the deadlock after a third round of consultations, on 3 May the King called new elections for 26 June: it was the first time there had been a need to repeat polls in democratic Spain.

The June 2016 electoral results did not help to break the deadlock. The path to incumbency, instead, saw the repetition of some of the pattern-breaking steps experienced during the previous months. What took place, in other terms, was a sort of ‘normalisation of the abnormal’. Thus the King led two separate rounds of consultations at the end of which he appointed Rajoy (leader of the first party) as candidate prime minister, first in July and then, again, in October. The new normal also included the rejection of Rajoy’s investiture by the Congress on 31 August and 3 September and, consequently, the restart of the two-month countdown to a third election. However, after being reappointed as candidate prime minister by the King, on 29 October Rajoy was able to win a parliamentary vote of confidence, a mere two days before the expiry of the 31 October deadline to call new elections. Rajoy obtained the support of a simple majority of MPs after failing to gain an absolute majority. Thus he became the third Spanish prime minister – after Calvo Sotelo in 1981 and Zapatero in 2008 – to have been voted in by parliament in the second rather than the first investiture vote.¹²

Overall, the process of installing a government after the 2015 elections broke several records. First, it took a very long time – ten months – during which the outgoing cabinet remained in office as caretaker; second, it needed two general elections; and, third, it entailed five failed investiture votes (two for Sánchez and three for Rajoy) before the PP leader was able to win one and form a minority government. Last, but certainly not least, it was the first time that a PP prime minister was elected with the abstention of the PSOE, that is of the main opposition force in parliament. In all the investiture votes since 1979, the second party had always voted against the first one. These features amounted to a radical breach in the traditional pattern of government formation and were accompanied by equally unprecedented events inside the political parties, as shown in the articles in this volume. Finally, while the outcome of this long process seemed to remain on traditional tracks, as it produced a minority single-party government, it actually entailed a radical element of novelty, since Rajoy passed the investiture vote with the abstention (instead of rejection) of the main opposition party, something that had never occurred before, not even in the consensual years after the end of the authoritarian regime.

Conclusions

During the first phase of the eurozone sovereign debt crisis, with its particularly severe consequences in Southern Europe, we noted that an ‘electoral epidemic’ was affecting the region as citizens manifested their political discontent at the polls (Bosco & Verney 2012). At the beginning of 2015, it appeared that the economic crisis in Southern Europe was beginning to recede. Nevertheless, the national parliamentary elections that took place over the next 18 months in Greece, Portugal and Spain registered even higher levels of public dissatisfaction, resulting in the spread of the electoral epidemic.

Following this latest electoral round, we find an electoral landscape in Greece (2012–15) and Spain (2015–16) characterised by frequent elections, increasing abstention, an astonishing plunge in loyalty to the traditional parties of government, extensive electoral volatility plus electoral fragmentation way above pre-crisis levels. In both these countries a ‘crisis-critical election’ took place – in May 2012 in Greece and December 2015 in Spain – producing a sharp and spectacular electoral realignment. The impact on the party systems was dramatic. Parliamentary breakthroughs by new contenders, changes in the identity of the three main parties, a major decline of the socialist parties and rising fragmentation radically reshaped the competitive arenas in these two South European states. The extent of change was striking in Spain and more extreme in Greece.

In comparison, electoral change in Portugal was more limited: above all, no ‘crisis-critical election’ occurred. Support for the two mainstream parties, levels of electoral volatility, and both electoral and party system fragmentation all remained more stable than in Spain and Greece. However, in Portugal as in the other two countries, voters’ perceptions of ideological distance indicated rising polarisation during the crisis era. Other party system trends, although manifested to a different extent, were also present in all three cases: the forces of the centre-right appeared electorally more resilient than the centre-left, the traditional communist left appeared to be merely surviving and the non-communist radical left experienced an electoral upsurge.

In 2015–16, it became apparent that the electoral epidemic had produced a government epidemic. During the early phase of the crisis, in 2011, all three incumbent governments had been ousted in Portugal, Greece and Spain. But at this point, in the Iberian states, government change came after elections (albeit early polls) and followed the standard pattern of alternation in power between the two main parties. Only Greece seemed an exceptional case, with its repeated departures from business as usual, including four different governments in 20 months. However, during the second phase of the crisis, in 2015–16, extraordinary occurrences in government formation became the new normal in the Iberian peninsula as well as Greece. It is striking that Portugal, which had seemed more immune to the electoral epidemic and less affected by party system change, nevertheless became part of the government epidemic. As we have shown, the route to government installation was characterised by unprecedented events and repeated elections. Meanwhile, government outcomes took these three South European states into uncharted territory. Moreover, the longevity of the governments formed could no longer be assumed. In short, in relation to incumbency, essentially nothing is as it was. From being almost boringly predictable in this sphere in recent decades, crisis-era Greece, Spain and Portugal became a source of surprises.

To conclude this article, the inevitable question is where Southern Europe is heading now. Perhaps the biggest effect of the electoral epidemic has been the disruption of the traditional patterns of bipolar competition that had characterised the three party systems for several decades. All three countries seem to have entered a system-changing era in which the logic of the electoral arithmetic points to coalition government. This even applies to Spain, where the current single-party minority administration was a last resort after repeated coalition formation failures. We have yet to see how the coalition outcomes of the government epidemic may impact on the electoral arena. The indications are strongest in Portugal, where the end of the inner party system and the revolutionary legitimation of the PCP as a potential government partner mark a systemic turning point. This development has changed for ever the traditional ‘imbalance between the left and the right in terms of the potential for coalition-building’ (Magalhães 2012, p. 311). It could – eventually – lead to a new type of Portuguese bipolarism based on alternative coalitions. The picture seems more blurred in Spain, where the competitive choices of the main parties are unclear at the time of writing, leaving the potential new government constellations an open question. Meanwhile in Greece coalitions based on the parties closest to the centre of the political spectrum have been succeeded by an alliance of apparent polar opposites originating from different ends of the political spectrum: new patterns of stable competition based on coalitions organised around the left–right axis have yet to emerge.

With considerable potential for a feedback loop from unusual government outcomes to electoral competition and party systems, changeability seems likely to remain a core characteristic of South European politics at least for the immediate future. It remains to be seen whether the destabilising developments making the path to incumbency arduous and the risk of government instability high will prove a case of regional exceptionalism or become a manifestation of a more standard European pattern.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Anna Bosco teaches at the University of Trieste and Susannah Verney at the National and Kapodistrian University of Athens. They are the Editors of South European Society and Politics and the related Routledge book series. Their previous co-edited books are Protest Elections and Challenger Parties: Italy and Greece in the Economic Crisis (2015), Southern Europe and the Financial Earthquake: Coping with the First Phase of the International Crisis (with Marina Costa Lobo, 2014) and Elections in Hard Times: Southern Europe, 2010–11 (2013) (all published by Routledge).

Acknowledgments

The authors would like to thank Ioannis Andreadis, André Freire, Eftichia Teperoglou and Emmanouil Tsatsanis for their help with survey data on Greece and Portugal.

Notes

1. The CDS changed its name to CDS–PP in 1995.

2. The effective number of parties can be calculated at the electoral or parliamentary level (either on votes or seats percentages) according to the formula of Laakso and Taagepera (1979). It ‘conveys information about fragmentation; for example, a figure of 4.14 tells us that the party system is “in effect” as fragmented as if there were 4.14 equal-sized parties’ (http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/Docts/effno.php).

3. For a very interesting discussion of critical elections and the issue of the (de)institutionalisation of party systems in Western Europe see Chiaramonte and Emanuele (2015).

4. Since 1987 the PCP has taken part in the elections allied with a small green party within the CDU (Coligação Democrática Unitária - Unitary Democratic Coalition).

5. It should be noted that our discussion will not address the question of the short-term service governments, usually headed by non-political figures, that were formed for a few weeks in order to hold elections.

6. For more on these two coalition governments, see Pridham and Verney (1991).

7. In the Spanish bicameral parliament it is the lower chamber that has the right to a vote of confidence in a government.

8. In Spain the prime minister may decide a premature dissolution of the parliament. Premature dissolution of the Cortes has occurred quite often and the only elections that have taken place after normal expiry of the legislature were those held in 2000, 2004, 2008 and 2015.

9. Under the Portuguese Constitution, the president of the republic appoints the prime minister in light of the electoral results and after consulting the parties represented in parliament. The nominated prime minister then submits their ministerial team to the president, who appoint the rest of the government. Next, the new executive must submit its programme to parliament and not have it rejected by a motion of censure (i.e. the government is invested by the parliament unless its programme is rejected by a majority). Such a rejection implies the government’s resignation. For more details on the rules on government formation in Portugal, see Fernandes and Jalali (2017).

10. In 1978 Nobre da Costa saw his cabinet rejected in similar circumstances. However, the 2015 occurrence was radically different, since Nobre da Costa had been appointed to preside over a cabinet that had been designed by President Eanes and did not represent any political party.

11. See Público, 24 November 2015, pp. 1–10; also, for an analysis of the evolution of the president’s real powers in Portugal see Fernandes and Jalali (2017).

12. Rajoy obtained the votes of the PP, Ciudadanos and Coalición Canaria (170) and could count on the abstention of 68 MPs of the PSOE. Only Rajoy in 2016 and Zapatero in 2008 received a simple majority of votes at their investiture. All other prime ministers, including Calvo Sotelo (whose investiture was interrupted by an attempted coup), received an absolute majority of votes at their investiture, even if destined to lead minority cabinets.