A few thoughts on Slovakia’s marriage referendum 2015

There is almost no circumstance that has not been covered first by Monty Python. Take for example the cricket coverage of Episode 20:

[Cut to fast bowler. He bowls the ball but the batsman makes no move whatsoever. T
he ball passes the off stump.]
Jim … and no shot at all. Extremely well not played there.
Peter Yes, beautifully not done anything about.
Brian A superb shot of no kind whatsoever

Yesterday most residents of Slovakia did nothing about the referendum on gay marriage, gay adoption and sex/euthanasia education and it failed quite badly as a result.

This is not much of a surprise.  As the graph shows, the citizens of Slovakia do not have a strong track record for voting in what they perceive to be minor elections, particularly the European Union, regional governments and referendums.  (Their turnout rates for national parliament, president and local elections are, by contrast, passable, at least from an American perspective, which isn’t saying too much. And turnout rates for all elections have remained quite stable for the last decade).skturn2015

The turnout for referendums is traditionally not as low as for European elections but as the graph below shows, it has hovered between 20% and 30% (closer to 20) with three exceptions: the 1997 NATO referendum which was exceptionally low–because of allegations of ballot tampering (not just ordinary tampering but rather the omission of an entire question by the Ministry of the Interior), the 1998 privatization referendum which occurred at the same time as the 1998 parliamentary election (and still did not reach the magic participation point of 50% even though overall election turnout was in the 80’s) and the 2004 European union referendum (which just barely managed 50% turnout even though most government  and opposition figures supported it and the stakes were extremely high). Into this environment stepped Slovakia’s Alliance for the Family (Aliancia za rodinu) with a referendum proposal (four questions, reduced to three by the Constitutional Court) for which it gathered over 400,000 signatures (almost all of them ultimately found to be valid, yielding well in excess of the 350,000 minimum).  The politics of gay marriage and gay rights in Slovakia I will not get into here, nor will I attempt to speculate whether the organizers thought they could actually overcome the 50% referendum hurdle or merely saw this in as a way to gain attention, identify supporters and force leaders to commit themselves one way or another.

Slovakia Referendum 2015 turnout

Source: http://www.politicaldatayearbook.com/Chart.aspx/38/Slovakia, download data here: http://bit.ly/pdyi_slovakiadata . See http://is.muni.cz/th/103226/fss_m/Diplomova_praca.txt for a good summary of Slovakia’s referendum history.

The actual turnout, 21.4%, ranks this referendum 5th out of the 8 in Slovakia’s post-1989 history, just below the 2010 reform referendum and just above referendums on privatization and early elections from more than a decade ago.  The actual share of “Yes” votes on the marriage question, 95.8% gave the marriage referendum the fourth highest positive share of the 14 valid referendum questions asked in Slovakia since 1989.  The adoption referendum received 94.3% support, for 6th place, and the sex/euthanasia referendum received 92.6% for 8th place.

Combining the low turnout with the high levels of support yields tells us the total number willing to turn out to support the topics of the referendum.  In the case of the 2015 referendum, this ranged from 0.202 supporters per registered voter for the marriage question (7th highest of the 14 questions), 0.198 for adoption (9th highest), and 0.193 (10th highest) for sex education.  The overall share of total supporters per voter was quite similar to that of the 6 reform questions introduced by the party Freedom and Solidarity (SaS) in 2010, which ranged between 0.161 and 0.218.  In this case, however, it is not clear that the real number of active supporters is quite so low: by contrast to the reform referendum of 2010 and other referendum questions, the 2015 referendum had to contend with worse weather and also with some degree of apathy about turning out to vote for a core policy demand (forbidding gay marriage) that had already found its way into Slovakia’s constitutional law earlier in the year.

Slovakia Referendum 2015 yes votesI find it useful to think about the turnout level in its geographical and temporal patterns as well.  As a battle of “turn-out” or “stay home” rather than “yes” versus “no, the spatial distribution of turnout may say more about regional support than does the actual vote margin.  The Slovak Statistical Office map shows a gentle gradient of support from highest levels in the north and east to much lower levels in the south and east, a near-perfect inversion of the 2001 census map (the only one I can easily find) of non-believers.


Turnout in the 2015 Referendum (darker is higher turnout).  Source: Slovak Statistical Office, http://www.statistics.sk

Subregions by share of non-believers (lighter=more believers). Source: http://www.unipo.sk/public/media/14066/peregrinus%20cracoviensis%2013.pdf

I took this opportunity to do something I’d never done before and compare levels of turnout by sub-regions and found some moderately surprising results:

Correlations between subregional voting patterns, by election in Slovakia 2010-2015

Correlations between subregional voting patterns, by election in Slovakia 2010-2015.  Red indicates more positive correlations, blue more negative correlations.  Depth of color indicates higher correlation strength.

Turnout patterns in the 2015 referendum most closely resemble the recent 2014 local elections and are quite different from those of the 2010 referendum–another turnout-based vote, but in this case one supported by the urban, secular SaS.  I haven’t taken time to dig any deeper, but I am struck at the correlation between that 2010 referendum, the 2012 parliamentary election, the 2014 European Parliament election and especially the 2014 presidential election.  This needs some more explanation.  I suspect that Vladimir Krivy would do it (or has done it already–I need to check.)


“The correlation is about this big.” (Vladimir Krivy knows all.)

Unlike the situation in Slovakia, a referendum forbidding gay marriage succeeded in 2013 in Croatia by a 2-to-1 margin with 38% turnout, but differences in strategy around the referendum law obscure a fairly high degree of underlying similarity.  Because Croatia does not have a minimum turnout requirement, opponents were forced to mobilize “no” voters rather than to encourage them to stay at home.  The resulting opposition boosted turnout above Slovakia’s levels, but not enough to sink the referendum.  The marginal turnout and significant-but-not-enormous share of “yes” voters meant that the clear win for Croatia’s referendum resulted from with only 24.96% of the total electorate, just 5 percentage points more than in Slovakia.  In a slightly different institutional environment, it is possible to imagine either of these referendums coming out quite differently.

Finally, a few haphazard notes about political parties and this referendum.  While it is nice to see that the Slovak National Party and the Party of the Hungarian Community have something in common, I am not sure that their shared support for this referendum is something that will bring them together on other issues or will lead to a more peaceful Slovakia.  I’m also not sure that this was the finest moment for many politicians except to the extent that it demonstrated their ability to dance around issues that they did not want to address directly (Question: “Are you voting for this referendum?” Answer: “I am voting in accord with my long personal history of supporting Slovaks and their families.” Question: “Yes, but…”  Answer: “Seeing that there are no other questions…”).  I have to admit being surprised by Radoslav Prochazka’s comment on the adoption question that “It is better for children to be in any household where people get along than in an orphange („podľa presvedčenia, že deťom je lepšie v akejkoľvek domácnosti, kde sa ľudia majú radi, ako v detskom domove“).  On a personal level–a place I have not usually gone with this blog–I would only prefer that he could have stopped at “along” and acknowledge the absolute rather than the relative value of a loving home, regardless of parents’ genders.

"Daddy, why is that man following us?"

“Daddy, why is that man following us?”

Austrian Parliamentary Election, 29 September 2013

In button form the Austrian flag looks a lot like a “Do Not Enter” sign, but this election at least that did not apply for new parties (of course the thing I am most interested in these days).  There was some significant volatility in Austria this year. In raw party terms, two new parties entered and one existing party left parliament.  There was a minimum of 14.5% shift in votes and a 16% shift in seats (less than in the last election, for which my figures are 17% of votes and 27% of seats, but much more than Austria’s overall average during the post-WWII period.)  The shift was of a different character, furthermore:  whereas nearly all of Austria’s volatility has in the recent past been among parliamentary parties (Mainwaring’s intra-system volatility, Tucker and Powell’s Type A volatility), this year nearly half of the total volatility (6.2% of 14.5) was related to new entries and exits (extra-system or Type B).  In the last election the extra-system volatility was nearly as high (5.9%) but it was outweighed by intra-system volatility (10.9%); in elections before 2008, extra system volatility was barely discernable (averaging only 1.3%). 

Particularly notable here is the emergence of the non-traditionally-named Team Stronach whose website ad looks like a blockbuster trailer:

Team Stronach Video

But of course the only interactive website anybody /really/ needs is the one that lets you look at the results yourself, the Political Data Yearbook: Interactive (he said fully aware that this post is simply a plug for that website:

German Parliamentary Elections, 22 September 2013

It’s seems to be election season in Europe–Norway last week and Germany this week, with Austria, Luxembourg and the Czech Republic coming up in the next two months (the latter two of which are somewhat premature).   No time (or expertise) to offer in-depth analysis of the German elections here, but I wanted to post some screenshots and a link to the Political Data Yearbook interactive: http://www.politicaldatayearbook.com/Chart.aspx/14/Germany

SDP did not t do that badly, at least by some metrics, and although this is being hailed as a “commanding election victory” and a “triumph” (which for CDU-CSU’s raw numbers, it certainly is), it looks a bit less impressive when seen in the context of the collapse of the FDP. Party death and birth, as always, are the interesting questions for me.  FDP seems to be the kind of party that can survive a single very bad election, especially if people quickly tire of what is being done in its absence from parliament, as they likely will.  But some support may also bleed off to another destination.  By my rough calculations, there was about a 5% net increase in voting for parties outside the mainstream, most of it to Alternative for Germany (among existing small parties there was no clear increase).  They appear to be betting now on poor economic results and consequences from bailouts to set themselves up and reap the dissatisfaction (not my analysis–that comes straight from National Public Radio (not the finest source on European domestic politics, but good enough, http://www.npr.org/templates/story/story.php?storyId=224101219).  So an inconclusive death and a not-quite birth.  The same whirlpools that tear Eastern European party patterns asunder with every election are not as powerful here and they face much stronger opposition.  But the currents still seem to be present…

Norwegian Parliamentary Elections, 19 September 2013

In the past I have reserved this blog for coverage of elections in Slovakia and the Czech Republic and for other elections where we see the emergence of new parties (one of my favorite topics).  In that sense, this post is the beginning of something new.  For the last two years, I have had the great good fortune to be allowed by the European Consortium for Political Research and Wiley Blackwell to help develop an online database derived from the rich 20 year store of data contained in the Political Data Yearbook associated with the European Journal of Political Research.  It is the intensive work on this that, for the most part, accounts for my neglect of this blog in the last year or so, and it occurred to me that I could remedy this not only with some coverage of the upcoming Czech election, but also with some cross-posting between my two projects.  For the moment, I simply want to announce that we have used the database to post the results of today’s Norwegian election.  There’s nothing unusual about posting results, but what sets the PDYi apart is its ability to put those results in graphic context: how does this election compare with the last 6?  Which parties did better or worse and by what margins.  The PDY shows this and allows for a considerable amount of choice of variables and modes of display.  A few samples below.  And once Norway’s cabinet is announced, we can post that as well to show what ministries have been held by what parties (and ages and genders) over time.  So take a look at the screenshots below and head to http://www.politicaldatayearbook.com/Chart.aspx/63/Norway for Norway or to http://www.politicaldatayearbook.com for all available data.


A sample line chart showing changes in party support over time.











 A sample stacked bar chart showing the overall success of the Norwegian right and center-right over time.