Sunday, September 29, 2013

Grand Coalitions and Austria's Radical Right?

Today’s Austrian election results can be viewed in a few different ways. On one hand, the “grand coalition” between the Socialist Party (SPO) and People’s Party (OVP)—the traditional mainstream parties of the left and right in Austrian politics—looks set to continue. On the other, support for the two parties dropped by a combined 4-5 percentage points, while three radical right parties—the Freedom Party (FPO), the Alliance for the Future of Austria (BZO), and the new Team Stronach—combined to win nearly one-third of the vote. In addition, the Greens had their best ever result, taking over 11% of the vote.

So what to make of these results? One interpretation is that they show a potential limitation with grand coalitions. For those not familiar, a grand coalition is a governing coalition between the two (or more) largest—and often ideologically and politically rival—parties. As such, it is a sort of unity government bridging left and right. Starting in 1986, grand coalitions between the SPO and OVP have been the most common government in Austria. As the chart below, suggests, this arrangement may well have been very costly for these two parties. The chart below shows the vote share of each major party (the radical right parties are combined into a single column), with the final column indicating whether the incumbent government was a grand coalition.

Grand coalition?

The chart shows a linkage between support for the radical right FPO, BZO, and Team Stronach, and the presence of grand coalitions. Prior to the first grand coalition, the SPO and OVP dominated Austrian politics, combining to win well over 80% of the vote. Now, that number is only slightly more than 50%. Periods of grand coalitions (1986-1999, and 2006-2013) have seen rising support for the radical right, while the radical right's support fell precipitously when it joined the government between 1999-2006 (the FPO was also in government between 1983-86).

Why would this happen? The answer, I believe, is that grand coalitions present a problem of choice to voters who are unhappy with the government and want a change. With both mainstream parties sharing power, there is nowhere to turn for voters but to the more extreme parts of the party system. In the case of Austria, it is either the radical right (FPO, BZO, or Stronach) or the Greens (who are extreme to many voters). 

Moreover, radical right parties often make populist claims a central part of their campaign strategy. In short, they claim to represent the "real people" of their country in opposition to collusive and corrupt elites. A grand coalition seems to embody that sort of collusion and corruption ideally; the mainstream political elites are ganging up to share the power of government, leaving ordinary people with no recourse! This undoubtedly makes radical right parties more attractive to voters. In the current eurozone crisis, where national leaders arrange bailouts and austerity agreements, such populist charges probably resonate even more.

(As an aside, this probably explains why support for the Greens is not related to grand coalitions. Presumably, they have gained support as a result of value change and shifting socio-economics).

Sustained periods of grand coalitions have possibly changed the landscape of party support in Austria. Below, I show the vote share of each party among those under 30 (blue columns) and those 60 and over (red columns). (Source). The results are pretty striking. The 60-and-over voters look like a traditional view of Austrian politics; the SPO and OVP get the most votes (about 65% combined), with the FPO a distant third. Young voters, however, are essentially split evenly between the SPO, OVP, FPO, and Greens. Is this the future of Austrian electoral politics? Maybe, maybe not. But it's clear that younger Austrian voters don't have the same attachment to the two big traditional parties that their parents (or grandparents) do. Many factors, such as value change or economic frustrations, could feed into that shifting support. But the fact that the two largest parties have governed together almost exclusively during this time has socialized younger voters to look either to the Greens (on the left) or the FPO (on the right) for an alternative to the incumbent.

This suggests that the SPO and, especially, the OVP may have to contemplate the possibility of pursuing governments with the parties further to their left and right to begin to reverse their electoral losses. But I wonder if those won't happen until the combined vote of the two parties falls below a majority.

Monday, September 23, 2013

About those Election Results...

Now that the German elections are over, we can start the post-election analysis season. And this truly becomes a silly season. Pundits will claim Angela Merkel and her CDU won because of her deft handling of the eurozone crisis, because of her steady leadership, because of her maternal characteristics, etc.

Before you get carried away, step back and remember a few points:

1. About 44 million Germans voted Sunday. Of those, about 20 million voted for the CDU/CSU. That means about 24 million voted for another party. Keep that in mind before you start up with sweeping claims about German voters. Many of those 24 million in fact voted for parties with quite different programs that Merkel.

2. People vote for all sorts of reasons. Did some Germans vote CDU for the reasons you want to claim? Probably. But most others probably didn't.

3. Don't confuse the effect with the cause. It's true that Merkel's re-election will have various policy consequences for Germany and the management of the eurozone crisis, among other things. Don't assume that people necessarily voted CDU for those reasons. One of those things about representative democracy is that you cannot choose which parts of the candidate/party you wish to support; you generally have to vote for the whole package. (I guess this can be taken as an argument in favor of open-list voting).

4. A voter is not the same as the electorate. A voter can decide how to vote. Electorates technically don't "decide" anything, though they represent the collection of millions of individual decisions.

In short, stick to the simple interpretations. Merkel and the CDU won because things are good, and voters tend to reward incumbents for good times and punish them for bad times. Any more specific claims could be true about some voters, but these sorts of claims require individual-level data and careful, controlled analysis. Not something you can get from aggregate election results and exit poll summaries.

Thoughts on the German Election

The German election results are in. The Christian Democratic Union (CDU) and its Bavarian sister party, the Christian Social Union (CSU) won a resounding victory, falling just short of an absolute majority. The opposition parties on the left--the Social Democratic Party (SPD), The Left, and the Greens--reached a cumulative total close to that of the the CDU/CSU. Nonetheless, Angela Merkel and the CDU are the day's big winners.

Meanwhile, the CDU's junior partner in the outgoing government--the Free Democratic Party (FDP)--failed to reach 5% of the vote and will not be a part of the next parliament due to Germany's 5% election threshold. As the FDP has been in every parliament in the modern federal republic--and has been in government more than any German party--this is a big result. Similarly, the newly-formed anti-euro Alternative for Germany (AfD) also came up just short of reaching the 5% threshold. Other smaller parties, including the Pirate Party, failed to make a splash.

49.4% (311)
30.5% (192)
10.2% (64)
10% (63)

In many respects, this was an election whose results are defined by the "almosts". The CDU almost won an absolute majority. The FDP and AfD almost surpassed the 5% threshold. These "almosts" are fairly consequential.

Germany's five-percent threshold law probably played a bigger role in this election than in any previous one. Previously, I described the consequences of Turkey's 10% threshold, and I even suggested how a recent election there would look different under Germany's more reasonable 5% threshold. But note that close to 15% of German voters will go unrepresented in this election due to the 5% threshold. This is a downside to the use of an election threshold. One consequence is that this was the most disproportional election in post-war Germany (courtesy of Armen Hakhverdian). Electoral disproportionality describes how much the distribution of seats in the legislature deviates from the distribution of votes (higher values mean greater deviance between votes & seats). Of course, this was an unusual circumstance with two parties finishing just shy of 5%. Nonetheless, this threshold-induced disproportionality helped to put the CDU/CSU so close to a parliamentary majority.

Understanding the Vote 
So what explains this result? The best single answer, I think, derives from economic voting. Angela Merkel seems to have bucked the trend that has seen incumbents in struggling eurozone countries lose elections decisively--successive Greek elections in 2009 and 2012, Ireland in 2011, Spain in 2011, and France in 2012. The difference is that Germany's economy is performing well. And most German voters are at least minimally aware of their good fortune relative to the rest of Europe. So it's no surprise that they rewarded the incumbent's party with another term in office for its management of the economy. Note the result from the chart. The top question asks whether "the economic situation" is "good" or "bad." Nearly three-quarters of German respondents (74%) say good, while just over a quarter (26%) say bad.

(Chart here)
(See also this analysis of unemployment and left-party support)

So far, so good. But the obvious question is why the FDP fared so badly when things were so good. Why did the FDP not also benefit from the good economy as a partner in the outgoing coalition government? One answer is that the head of government's party is held accountable more than junior coalition partners for the government's performance. This claim has support in the economic voting literature. This paper by Fisher & Hobolt (paywalled) shows that the prime minister's party is sanctioned much more than junior coalition parties for retrospective assessments. Still, such a claim does not explain why the FDP's support collapsed from 2009. That has to be considered a weakness of the economic voting model in this case.

What does this election tell us about German politics?

Do these results represent any sort of "sea change" in German politics? Had the AfD crossed 5%, it would be far more tempting to say yes. Instead, the results suggest (to me) politics as normal. Three big questions might be asked:
  • Can the SPD win an election again?
  • Is the FDP finished?
  • Is the a start for the AfD or a flash in the pan?
Regarding the first question, it depends on what "win" means. The left seems to have reached a three-party equilibrium of the SPD, Greens, and the Left. The Greens seem to attract younger, more affluent, urban voters, while the Left peels off economically marginalized voters in the former East Germany. If one looks at Sweden, for example, a sort of reverse situation exists in which multiple right-wing parties can combine to govern in spite of the fact that the Swedish Social Democrats always "win" each election. So the SPD does not need to get the most votes necessarily to govern if the left as a whole can command a majority. The long-term strategic dilemma is its relationship with the Left. As long as it refuses to govern with the Left, it will be hard for an SPD/Green coalition to win a majority--and that leaves it with few options besides an occasional grand coalition with the CDU as a junior partner, which may well be the "worst of both worlds" from an electoral point of view. But SPD leadership worries (and for good reason) that governing with the Left could damage its credibility among mainstream voters.

Considering the FDP, the answer is surely "no". The last two elections have seen the FDP swing between extremes of over 14% and under 5%. In truth, its "natural" level of support probably remains around 7-8%. Because Germany is federal, the FDP still has a base in many Land parliaments from which to begin rebuilding itself for the next federal election.

And what about the AfD? The following chart asks voters whether they decided to vote for the party out of conviction (left) or disappointment (right). Only for the AfD did a majority of voters choose the party due to disappointment, while it was close to evenly-split for voters of The Left. That may tell us something about the future of the AfD: only 37% of its voters were genuinely supportive of the party by their own report. To me, this suggests a pretty grim future for the party. It did not attract many convinced supporters. And, unlike the FDP, it has no base in regional parliaments upon which to build/sustain itself. This is not to say that there is not room in German politics for a party like the AfD (I think there definitely is), but it seems pretty hard to build on failure in this manner.

(Chart here)

Wednesday, September 18, 2013

Pre-Electoral Coalitions and Voter Turnout: Understanding How PR Affects Turnout

This blog is about my most recent published article, forthcoming in Party Politics and available online here (paywalled) or here (not paywalled) in manuscript format.

Proportional representation (PR) electoral systems appear to generate contradictory effects on voter turnout. What do I mean by "contradictory effects"?

PR systems appear to increase turnout directly. Or, to put it more accurately, single-member districts (SMD) systems discourage turnout. How? PR systems make the act of voting more worthwhile for all voters because votes for small parties are not “wasted” as they are in SMD systems. In other words, supporters of smaller parties do not face the strategic choice of either "wasting" their vote on a smaller party or voting "tactically" for a large, less-preferred parties. More importantly, SMD systems make voting much less important in certain districts than in others. Think about what happens if you live in an electorally uncompetitive legislative district (like all of the Congressional districts in the City of Chicago, where I live, or many others in the US or other SMD systems). Is there really much point in voting when the outcome is a foregone conclusion in that district? (Hint: the Democrat will win where I live!). Not really. Of course, there are other reasons to vote (such as to affirm democracy or to fulfill a duty of citizenship). But a major reason for voting is to select who will represent us, and that becomes relatively pointless in an uncompetitive district. Based just on this direct effect, we would expect turnout to be higher in elections conducted under PR electoral rules.

However, PR systems also appear to decrease turnout indirectly. How? The answer is by leading to the development of multiparty systems. That is usually thought to be one of the virtues of PR. However, elections in most PR systems become less decisive because no single party will actually win a majority of the parliamentary seats. As a result, there will need to be some sort of bargaining among party leaders to form a government (given that, somehow or another, a parliamentary government needs to pass a confidence vote to take office), and voters may not be able to anticipate the correct result or influence it. For example, consider the situation after the 2010 Swedish general elections.

Seat %
Social Democratic
Liberal People’s
Sweden Democrats
Christian Democrats

Without knowing anything about Swedish politics, could you guess the eventual governing coalition that formed? Probably not! And this confusion, along with a sense that your vote doesn’t matter as much because the party elites will decide the government among themselves anyway, might convince you that it’s not worth the time to vote.

But, in fact, Swedish voters did have a pretty good idea about what government might form. Why? Because four of the parties—the Moderate Party, Liberal People’s Party, Center Party, and Christian Democrats—had made a public commitment to govern together before the election. This public commitment is called a pre-electoral coalition (PEC). Party leaders form PECs in order to send a public commitment about their plans to govern together, and other research has shown that these commitments are almost always honored (meaning they are credible to voters). One effect of a PEC is that it makes the forthcoming election clearer: now a voter knows that a vote for any of the parties in that PEC will be a vote for the coalition itself. In the Swedish case, voters now knew it was a choice between the four-party coalition or a Social Democrat-led government, which made the election much more like a choice between two rival alliances.

In this new article, I take advantage of the fact that PECs are formed before some elections but not before others to examine what effect they have on voter turnout. If the arguments about PR and turnout are correct, then the presence of a PEC should increase turnout. Why? Because the potential choices in the election become clearer.

The results of the analysis I conduct in the paper show that turnout is about 1.5-2.0 percentage points higher when there is a PEC than when there is not. This shows that decisiveness matters: people are more likely to make the effort to vote when they can see how their vote will determine the government that takes power. This helps to clarify our understanding of when and why people vote.

These results highlight a fundamental tradeoff with electoral systems. Among other virtues, PR electoral systems increase turnout directly by making voting more attractive. But they indirectly may discourage voting by making elections less decisive (or clear) to voters. Variation in the existence of another factor that increases the clarity of elections (PECs) helps to solidify this interpretation.

In short, there is no “perfect” political institution. Every particular institutional design comes with its own tradeoffs. In some cases, the advantages may clearly outweigh the disadvantages. But the disadvantages will still be there. As political scientists, we can play a useful role in public debates by identifying and explaining those tradeoffs.