Spaniards head to the polls again on June 26 — just six months after the last elections. Here’s why: As predicted here in the Monkey Cage, the four main parties — the conservative PP, the socialist PSOE, the liberal Ciudadanos and the radical-left Podemos — failed to come together to form a government, even though some of them (mainly PSOE and Ciudadanos) certainly tried. Since the parties weren’t able to form a coalition government, Spain will hold a new election.
Is Spain wasting time and money to repeat an election that led to a deadlocked government? I looked at some historical precedents, not only in Spain but also elsewhere in Europe. Similar “snap” elections really didn’t change much in terms of electoral preferences, number of parties or government alternatives. Here’s what I found:
1. In Turkey, expect the unexpected.
Like Spain, Turkey needed a repeat election within a few months. Turkey’s June 2015 elections left the governing Justice and Development Party (AKP) short of an absolute majority. Those elections also returned four main ideologically distinct parties: the Kurdish HDP, the socialist CHP, the nationalist MHP, and the conservative AKP. When AKP was unable to form a surplus majority government, the country held new elections in November 2015. AKP won outright and launched its fourth single-party government since 2002.
2. Spain has a history of voting paralysis — and collapsed governments.
To find a Spanish precedent for consecutive elections, we need to go back a whole century. Between February 1918 and December 1920, Spaniards had to turn out and vote every year. The parallels with the current political situation are striking.
Like the 2015 elections, Spain’s 1918 elections returned a particularly fragmented parliament. Similarly, the 1918 elections also put an end to the bipartisan split that had pitted Liberal Conservatives (PLC) against Liberals (PL) and dominated Spanish politics since the Bourbon Restoration in 1888. But actually forming a lasting government turned out to be extremely difficult. Up to six different PL-led cabinets, two of them national unity governments and one semi-authoritarian, followed. Five of them lasted no more than three months.
Parliament ultimately dissolved in an attempt to end the political instability. But the June 1919 elections returned an even more fragmented legislature. The PLC managed to form two short-term cabinets, both of them single-party minority governments. A national unity government ruled for five months in between the two.
Like current Spanish Prime Minister Mariano Rajoy, Eduardo Dato — the conservative leader at the time — thought he could increase his support with another round of elections. The results of the December 1920 elections did not change the political panorama so much. The PLC did manage to increase its number of seats, but divisive politics and cabinet instability continued to be the norm. Three years later, democracy definitively collapsed with a military coup orchestrated with the king’s approval.
3. In other countries, voters needed more time to change their votes.
Seven other countries held repeat elections within a six-month span over the 1848 to 2015 period. For some countries, it happened multiple times. The data from these elections, shown in Figure 1 here, reveal that voters tended to keep with the same vote as before.
The graphic shows the level of aggregate change in voters’ partisan preferences between two elections. Political scientists refer to this as “electoral volatility.” With just three exceptions, snap elections showed far lower electoral volatility. And in 10 out of these 13 elections, fewer than 9 percent of voters cast their votes for a different party. Perhaps because the short time period between one election and the next meant few real changes in party messages and promises, most voters decided to remain faithful to the partisan choices they had made less than six months earlier.
4. Repeat elections reduce fragmentation.
But repeating elections did tend to concentrate the playing field and lower the number of relevant parties in the electorate. The data for these same elections show reduced electoral fragmentation. So in Figure 2, below, only Germany (1932), Greece (2015) and Moldova (2010) saw a higher number of relevant political players after a follow-up election. And in seven out of 13 of these elections, the identity of the parties remained exactly the same. In the other five elections, the number of new parties never reached more than two.
Here’s what’s happening: Well aware that they don’t have enough time to get voters to change their minds, parties strategize their best move. They seek to maximize their electoral support by forging alliances that will improve their previous electoral returns. The electoral coalition recently forged in Spain by Podemos with the communist United Left (IU) is a clear example.
5. How much actually changes in a follow-up election?
Figure 3, below, looks at the patterns of government continuity in consecutive elections. I call this systemic closure — this means how predictable the composition of the government is after elections. The data for these same elections over this same period show that stability seems to be the norm. For nine out of 13 repeated elections, the same party managed to win.
In cases where the opposition managed to achieve victory in the first election, it retained the government in the second (repeated) election. This was the case in Denmark (1920), the United Kingdom (1974) and Greece (1933, 1990, 2012 and 2015). If the government managed to retain power in the first election, it immediately lost it in the second – this happened in Denmark (1953) and Moldova (2010). The only exceptions were Ireland, due to the Society of the Gaels’ (CnaG) governmental monopoly between 1923 and 1932, and Portugal, where the opposition won every election.
So what does all this mean for Spain’s June 2016 elections?
This analysis of repeat elections in eight different European democratic countries since 1848 can help us to understand what’s likely to happen in the coming Spanish elections. And the simple answer is: not much.
First of all, other than blaming one party or the other for blocking the process of government formation, voters do not have enough reasons to change their partisan choice. Second, the number of relevant electoral players is even less likely to change. The Spanish party system will continue to be represented by four main political players (PP, PSOE, Ciudadanos and the Podemos-IU coalition). The distribution of these four political forces along two different blocks, one on the right (PP and Ciudadanos) and another on the left (PSOE, Podemos, IU), will depend on the “sorpasso” (i.e., overtaking) of PSOE by the Podemos-IU coalition, as some opinion polls suggest.
But taking into consideration not only the Turkish example but also the tendency for a party in government for just one legislature to retain the government — as shown in Figure 3 — a victory of the leftist opposition would certainly constitute a huge surprise.
It’s even less likely that Spain will face a third election. In 168 years of democratic history in Europe, countries had to go to the polls a third time within three consecutive years (or less) only six times (i.e., Denmark, Greece, Ireland, Moldova, Spain and the U.K.). For this reason, we should expect Spain to be able to form a new government after these elections. How long this might take, and whether current party leaders will be at the head, remain to be seen.