“December Agreement”–Swedish election unsnapped

Sweden was supposed to have a snap election in March; it was one of my three examples of ways in which a cabinet can be terminated in a parliamentary system just over three weeks ago. Now Sweden offers an example of how a “snap” election can be called, and then called off. In fact, I did not know this was possible. I can’t think of a similar case offhand.

Prime Minister Stefan Löfven, of the Social Democrats, and the four center-right opposition parties, along with the Green Party, have struck a deal to allow Löfven’s government to survive, and thus there will be no election in March. The anti-immigrant Sweden Democrats and the Left Party were not part of the “crisis talks” that produced the deal.

The BBC indicates that,

Under the deal, Mr Lofven will follow the opposition’s budget next year, although he can make some changes.

However, after 2015, the agreement “commits the opposition to abstain from voting against the government’s budget proposals” and “co-ordinates the parties’ polices on pensions, defence and energy issues.”

Löfven’s cabinet is a minority in parliament, hence both the initial budget defeat and election call, and the motivation for a deal.

The deal lasts till 2022. This is all quite extraordinary.

The ways a cabinet can be terminated

There is a literature in political science on government (cabinet) termination in parliamentary democracies. This is not a review of that literature. Rather, it is an accounting of three recent cases that illustrate different ways that a government in a parliamentary democracy can end.

We can have a party based on an electoral majority that seeks a fresh mandate. That is, a party–or bloc of parties running together in the election– that has won a majority of seats, but, for strategic reasons, decides to hold an early election. Exhibit: Japan. Prime Minister Shinzo Abe (Liberal Democratic Party) announced a snap election on 18 November, despite about two years remaining on the term and a comfortable majority won in the 2012 election. The election will be 14 December.

We can have a coalition government, formed by bargaining among multiple parties after an election, which holds a majority of seats. The parties might have a falling out over one or more policy issues, and the parties break up the government rather than resolve their differences within it. Exhibit: Israel. On 2 December, Prme Minister Benjamin Netanyahu fired two ministers, each of whom heads a party that was in his coalition: Finance Minister Yair Lapid (of the Yesh Atid party) and Justice Minister Tzipi Livni (of HaTnua). The firing, which meant the breakup of the coalition, followed weeks of rancor over the budget and a bill backed particularly by another coalition partner, the Jewish Home party, that critics (including the ceremonial president) said would elevate the state’s Jewish character over its democratic character. New elections will be 17 March.

We can have a minority cabinet, in which the party or parties holding ministerial positions do not hold a majority of parliamentary seats, with no available backing from among the non-governmental parties in parliament, and which fails to get the support needed to pass its budget. Exhibit: Sweden. On 3 December, the minority center-left government of Prime Minister Stefan Lofven lost a budget vote, 153-182, and announced there would be a snap election. The wild card here was the Sweden Democrats, a far-right or “populist” party that neither Lofven nor the center-right opposition wanted to be seen to deal with following elections that were held less than three months ago. When the center-right would not back the budget, the government effectively lost its right to continue governing. A new election is expected to be set for 22 March.

One could say that these were listed not only in chronological order, but in reverse order of “necessity”. There was no reason why Japan needed an early election–its government has a solid majority. The Israeli election call is also not exactly necessary–the parties whose leaders Netanyahu fired were not in anything like open rebellion even if they were in policy disagreement (which is, after all, natural in coalitions). It was, however, a coalition that the PM himself clearly never wanted, and was forced upon him by the election results in early 2013 and the bargaining stances of other parties (specifically, Yesh Atid and Jewish Home). Polls show the potential of a much stronger right-wing bloc in a new election. So, it is an opportunistic call, but arguably less unnecessary than Japan’s. The Swedish situation, on the other hand, is one of real deadlock.

Of all these countries, the one that has the lowest tendency towards early elections is Sweden, even though minority governments are quite common there. Usually, however, they have had a fairly reliable “outside” party to back them on budgets or other confidence matters. Israel has frequent early elections–although this one will be earlier than any in years–and usually has oversized governments (meaning containing more parties than actually needed to have a parliamentary majority–such as the just-collapsed one). Japan usually has electoral-majority governments, but has had many early elections, including the famous one of 2005, also called by a government that had a secure election-based majority, but wanted (and got) a bigger one to push through reforms blocked by intra-party resistance.

All these great examples of early elections in parliamentary systems, and it wasn’t even my teaching quarter for any of my comparative democracy courses.

Sweden’s 2014 election

In Sweden’s general election on 14 September, the bloc of center-left parties headed by the Social Democrats won 43.7% of the vote, with the incumbent center-right parties reduced to a combined 35.3%. The anti-immigrant Sweden Democrats have around 12%, or roughly a doubling of their support from 2010.

Stefan Löfven, the Social Democrat leader, has ruled out cooperation with the Sweden Democrats, as has the outgoing center-right leadership. That means that the statement I have heard and read in various media that the Sweden Democrats “hold the balance of power” is evidently false. In fact, center-right PM Fredrik Reinfeldt has already resigned.

The Guardian reports:

Löfven hinted instead at deals with the two smaller parties in the country’s rightwing alliance, whose combined 11.5% share would bring them close to a majority. “I want to say that the hand is extended to other democratic parties,” he said. “Our country is too small for conflict.”

I do not know enough about Swedish politics to assess this. Maybe someone can help. The two major blocs are often reported as “alliances”, in the sense of signaling in elections their intent to govern together, given a favorable election result. Normally, I’d expect pre-election alliances to imply also going into opposition as a bloc. This sort of election result can change things, but how readily would parties within an alliance in Sweden break and extend support to (or even enter a cabinet of) a government of the other bloc?

A related question is, to what extent (if at all) do the parties within an alliance cooperate in the elections themselves?

A final note, also from the Guardian:

Reinfeldt’s minority government benefited from the tacit support of the far right, whose MPs voted in favour of an overwhelming majority of their measures. But it has always refused any formal cooperation.

Presumably even such tacit cooperation is less an option for Löfven, and appears in any case to be ruled out by his public statements.

And a final final note: Löfven had not previously been elected to public office; this is quite unusual for prime ministers in parliamentary systems, especially long-established ones. In fact, just 21 of 377 (5.6%) of PMs in parliamentary democracies have no prior experience as an elected national or regional MP, or regional or municipal executive, according to the Samuels and Shugart dataset.*

_________
* Or 24 of 391 (6.1%) have no parliamentary experience. (There are some missing data on some of the other experience variables.) For democracies older than the median parliamentary system in the dataset (18 years), only 6 of 203 (2.96%) PMs lack prior electoral experience.