My own comments here have less to do with Belarus than with what my experience in Belarus says about conditions necessary for genuinely free and fair elections.  One of the best meditations of this question can be found in Andreas Schedler’s magnificent 2002 Journal of Democracy article, “The Menu of Manipulation” (posted here –one hopes with permission– by the Netherlands Institute for Multiparty Democracy).  Schedler suggests that there is an almost infinite variety of mechanisms by which those in power can artificially inflate their own vote totals:  they can artificially add supportive voters (carousel voting) or ballots (ballot-box stuffing); they can artificially subtract rival candidates (ballot or media restrictions, or divide et impera through paper candidates) or rival voters (voter list manipulation or voter intimidation), and, of course, they can simply change the numbers themselves at some point during or after the counting process (which Lukashenka, in what may have been an attempt at humor, claimed to have done in 2006 to bring down his margin of victory).  Although OSCE observers saw nearly all of these methods in Belarus in 2010, the main problems arose during the counting process.

There is something striking about the counting process in Belarus.  The Ukrainian and Mozambican counts I have witnessed were loud and long and sometimes chaotic, but they produced a reliable result because many different voices contributed to the outcome, even in precincts where one candidate won over three-fourths of the votes.  In the Belarusian elections, by contrast the counting is quick and quiet, over almost before it starts.  This lets electoral observers come home earlier than usual, but they have to spend the rest of the night writing up reports.  In some things speed is a disadvantage; a fast vote count is about as useful as a fast clock.

So what to do?  External incentives may help those in power to allow a freer electoral process and a fair count, but in a place such as Belarus, the fast and unwatched count may be so habitual at the local level that anything but a change in the incentive structures of electoral management or direct monitoring from above may not be enough to change customary practice.  External observers can help, but their lightning checks are hardly enough to change anything (easier just to wait until they go away) and disregard for them has frequently manifested itself even during the counting process.

Does the presence of external observers do anything, then, other than generate data for reports?  Does it have any effect?  An interesting test of observers’ limits and capacities can be found in the OSCE’s unusual decision to post stationary observers who remained at the same polling place for three days of early voting.  My informal discussions with fellow observers suggest that many of those who did stay in the same place ended up building a degree of a rapport with the electoral committee of the polling place.  The rapport, in turn, seems to have resulted in more open and transparent processes–particularly during the vote count–than in those polling places where observers dropped in with little prior contact.  Having built up some degree of trust, the counters and observers were reluctant to wreck their tentative rapport and sought ways toward a mutually acceptable process.

But stationary observation is more a metaphor than a model.  There is neither the money nor the will (nor the tolerance of locals) to allow for the thousands of external observers it would take to build relationships in each polling station.  Only local observers can perform that role–and I have seen them do it with great skill and enthusiasm in Ukraine and Mozambique–but they too face limits.  While external observers may lack the relationships that could bring about a better process at the local level, but they are at least protected from political consequences of their observations.  Domestic observers, by contrast, have the necessary relationships but they may face power asymmetries that prevent them doing their jobs (or even from wanting to try).  In small communities with dense social networks, observers might be able to force a reasonable process, but they usually opt not do so because the costs of becoming involved in electoral antagonisms are higher than the benefits, especially where state-run organizations control the economic livelihoods of local residents.  There are greater chances for lasting effect in more anonymous urban communities where there are fewer social- and employment-related threats to prevent meaningful election day oversight.  The strong police effort in Minsk on election night may have been a way to send a message about the dangers of getting involved to those who faced fewer community constraints.

It is thus difficult for observers to have a direct effect on the conduct of local elections without a particular combination of connection and symmetry: the observers and the observed must have a relationship with one another, and both sides must have access to some form of power.  This combination is hard to maintain without a strong civil society or significant opposition parties that might plausibly gain control of regional or national governments.   The process of democratic competition thus contributes directly to the impartiality of local-level electoral administration, and the absence of one or the other can begin a vicious cycle that further grinds down both.  The dynamics of observer-observed relationships are just another reason for the cycle of gradual decline and sharp recovery exhibited by a number of postcommunist countries: since there is little opportunity for gradual improvement, things tend to get worse until something snaps yielding a new balance of power that may for a time restore a degree of fairness to the electoral process (as it appears to have done in Ukraine).

Belarus has not yet reached that point and I have found little compelling evidence that would allow a meaningful guess about when–or even if–it will.  As Timur Kuran pointed out nearly two decades ago in his “Now Out of Never“ the answer lies in a hidden landscape of concealed preferences that we will only glimpse if change actually occurs.

Viorel--the best partner one could hope for--makes friends in Mstislavl

Richard Swales is from England, but lives in Kosice where he is married and runs Jazykova skola Start (www.jazykova.sk), a school for Slovaks who want to learn English.

Richard and I have never met in person but he been such a consistent source of astute comments that I have asked him here to elaborate on an email he sent to me about preference voting in the most recent election in Slovakia and he was kind enough to do so.

In Slovak elections, the system used is the party list system. This means that each party submits a list of 150 candidates, and the voters choose the party they prefer. All parties which have at least 5 percent of the votes get seats in parliament in proportion to their vote share. For this purpose a number called the REN– republic electoral number–is calculated; this time it was 14,088 and parties got one seat for every 14,088 votes they got.  The question this post looks at is which of the 150 candidates get those seats. In some countries, the order in which the candidates are printed on the list is used, so if a party gets 15 seats, the top 15 candidates are elected (this is called a closed list system). In other countries, the voters also get vote for individual candidates to determine the order (this is called an open list system).

In Slovakia, each voter may choose to vote for the party only, or may choose to circle the names of up to four candidates on the list. Seats are awarded to candidates in order of the number of voters who circled them, but there is a catch. To get a seat in this way, a candidate must be circled by at least 3 percent of his or her party’s voters, otherwise the party-determined ordering of the list takes over. This can be called a semi-open list system.

What I want to look at is to what extent members of parliament for the different parties owe their position in parliament to having a constituency of supporters that circled them, and to what extent do they owe their position to the party placing them high on the list and being favoured internally in their party.

KDH 15 seats: 10 elected through circling, 5 elected based on their list positions.
Of those elected through circling, 1 was not in the top 15 on the original list (and so wouldn’t be in parliament if no circling had taken place) Of those elected based on list positions, none were in the top fifteen on circling (and so all five wouldn’t be in parliament if there were no 3 percent cut-off for circling candidates).

I don’t agree with the standard way of analysing elections of this type, by which the above would be taken to mean that the public were only able to change the KDH list order by one seat. For example Durkovsky (13th on the list, 8th in terms of circing) would not be in parliament if his voters had circled other particular other candidates and got them over the 3 percent mark. An alternative explanation is that only small numbers of voters disagreed with the ordering of the KDH list anyway, which of course featured the best known people at the top. Only seven candidates on the list got more circles than REN so could be said to have pulled their weight by bringing in as many supporters as needed to elect one member, so it’s difficult to say that candidates scoring less than 3 percent (6480 votes, less than half one REN) were unfairly excluded by the 3 percent rule.

Most-Hid, 14 seats
All members were elected by circling, and a further two members had more than 3 percent but were outside parliament, meaning that the list ordering had no effect (uniquely among party leaders, if Bela Bugar had got no circles he would be out of parliament even though he was first on the list). A total of 4 of the members elected were outside the original 14 on the list, including 3 OKS members (Most-Hid gave some spaces on their list to this smaller party, in the hope of pooling votes). The number of circles received by OKS leader Peter Zajac was just over the REN, which suggests that with 4 members of parliament OKS is by far the most over-represented party in parliament, and Most-Hid the most under-represented. Most-Hid voters were the most frequent users of the right to circle candidates (82.88 percent circled at least one candidate). This is consistent with past results for SMK, probably because in parties based on ethnic politics, voters have more need to use the circles to express more traditional right-left or conservative-liberal preferences.

SaS 22 seats: 10 through circling, 12 through list positions.
Of the 10 circled, 4 were not high enough on the list to be elected anyway. These were the Obycajni ludia group who were given the bottom four places on the SaS list. Their leader’s 38000 votes is almost 3 RENs meaning there is a stronger suggestion than in the case of OKS that these were not mainly at the expense of the party hosting them on the list. Of the 12 elected through list positions, 8 had high enough support that they would be elected anyway through circling if the 3 percent rule didn’t exist. SaS voters were the least likely to circle candidates (68.09 percent used this option), possibly reflecting that as a party of newcomers they had fewer famous candidates.

SDKU 28 seats: 10 through circling, 18 through list positions
One circled candidate was not high enough on the original list to be elected without the circles. Of the other 18, 13 candidates were circled by enough people to be in if the three percent rule didn’t exist and only circling was used.

SNS 9 seats: 8 through circling, 1 through list position.
One of the eight circled was outside the top nine on the original list. The one candidate elected through list position would not have had enough circles to be elected if the three percent rule didn’t exist.

Smer 62 seats: 9 through circling, 53 through list position.
The 9 candidates elected through circling were all from the top 62 of the list (in fact all from the top 10). There were 13 people lower down the list who had higher numbers of circles than 13 who got in through list position. It is worth pointing out that without the 3 percent cutoff a candidate would require 1,461 circles, to get into parliament so 0,16 percent of SMER voters or less than a tenth of the REN. I must say that I support the use of the 3 percent rule or a similar rule (possibly based on a fraction of the REN as in the Benelux countries, so it would apply more consistently across parties) as numbers of circles like 1100 or 1300 are just about who has an extra 200 people living in their home village.

Totals

61 candidates were elected through circling of whom 11 would not have been high enough on the original list ordering. in addition to those 11, a further ten candidates (the Most ones) are also heavily dependent on circling as the list order didn’t need to be used at any point

89 candidate were elected through list positions, of those 28 would not have had enough circles if the 3 percent rule did not exist.

Conclusion, the circling and the list orders were broadly consistent with each other, but there are a large number of members of parliament who owe their position to the party and its internal politics more than the use of circling.

-Richard Swales

What were the most surprising moments for you in the election results? Was there anything completely unexpected?

I did not expect how inaccurate the final week of polling numbers, and I was especially surprised by the difference between the exit polls and the final results: 6-7% difference between polls and reality for Smer is quite shocking.  Even exit polls are usually are not particularly good at predicting final results, but for both FOCUS and MVK to be so far off and in the same direction suggests a need to rethink polling and especially exit-polling methods, something that has occurred in a variety of countries including my own.  Fortunately, this particular surprising had little consequence except for those prone to extremes of euphoria or melancholy and by 02:00 we could more or less predict the actual composition of the next coalition.  As for the results themselves, I the only real surprise was in the relative percentages obtained by Most-Hid and MKP-SMK, but since I figured that I had no way of knowing how those would come out, there is no result that would not have surprised me in one way or another.

Beyond the election results themselves, the final days of the campaign produced some remarkable moments:  HZDS giving out flour as an election enticement; Fico so concerned about a minor competitor as to brandish a “Don’t vote SDL” sticker; the moment of 00:42 on election night when the updated election returns put HZDS below 5.0 for the first time ever in its history; and my sudden realization around 02:00 that Slovakia would have its first female prime minister, news which my 6-year old daughter greeted with great enthusiasm.

The election turnout was expected to be very low, about 50 percent, yet the actual number hits 59 percent. What’s the reason for this unexpected interest of voters in the elections, in your opinion?

Well the problem is in the expectations rather than the interest.  In my research for our last interview, I found that voting intentions were not running behind 2006 and that in other elections there had been a slight uptick in turnout from the mid-2000s to the late-2000s and so there was few grounds for expecting a sharp decline.  I think that a small share of turnout increase may be related to the emergence of several new parties which perhaps offered some motivation for turning out:  SaS, Most-Hid and SDL between them attracted more than 20% of the vote; it is not to big a stretch to think that without plausible new parties (a condition that prevailed in 2006) some of these voters might have stayed home.

The election results of HZDS has encouraged political analysts and opposition politicians to talk about the ‘end of the Meciar era’. What does the fact that they were left outside the parliament mean for the party’s future?

The day after the 2006 election I plotted out trendlines for the various parties’ parliamentary delegations and found that the trendline for HZDS pointed precisely at 0 in 2010.  I turned this into provocative graph for my blog, but Inever ever expected it to be accurate.  Until about two months ago I figured the party would have enough reserve strength to scrape over the barrier.

Now that HZDS is out of parliament, it will be even harder to reverse that trend (though I don’t expect it to achieve -15 seats predicted by the trendline for 2014).  Three forces work strongly against HZDS in the coming four years: history, demographics, organization.

• History is not a causal factor but something to take a close look at to determine causes.  With parties almost never return to parliament after dropping below the 5% threshold.  Once out of parliament, parties tend to be forgotten and sink even further or disappear altogether as the examples of ZRS, SOP, ANO, and KSS indicate.  In Slovakia the only major exception I can think of is SNS, which re-formed after its split in 2000 and manged to recover its initial electorate.  In the neighboring Czech Republic the Party of Greens also managed to re-enter parliament after a long hiatus but only as a completely reorganized organization with a new leadership.  HZDS lacks similar quick fixes to problems with its electorate and demographics.
• Demographics plays a major role in the sense that HZDS already had the oldest electorate in Slovakia’s politics and it was aging at about 1 year per year.  Even if all of its voters stayed loyal–something that obviously did not happen between 2006 and the present–it would still have fewer voters than it does now (and despite the picture of Meciar with a laptop, a party that distributes flour on election-day does not show strong signs of being able to reach out to younger voters).  HZDS also has the oldest leader of any of Slovakia’s major parties.  Meciar will be 72 in 2014 and he is already not the active campaigner and public presence that he once was.
• HZDS organization will also work against the party’s recovery.  A party with 4.4% might be able to recover to 5%–and both the Party of the Hungarian Coalition in Slovakia and the Christian Democrats in the Czech Republic will be trying to demonstrate their resilience–but not with the same leadership–something that the leadership of MKP-SMK and KDU-CSL recognized in their immediate resignations.   The problem for HZDS is that Meciar has systematically created a party that cannot live without him.  Since every HZDS leadership challenge has ended up with the departure or expulsion of the challenger there is really nobody left in the party.   HZDS and Meciar are inseparable and for that very reason both appear to be without a political future.  As an afterthought, I suppose it is theoretically possible to envision HZDS surviving in some weak form as part of an electoral coalition whose members might share 7%, but right now it is not clear what party would willing to join with it in that effort, and if, as I suspect, HZDS’s poll numbers begin to fall consistently below 3% it would not become a particularly appealing partner.

What is behind the very high result of Smer? Why, on the other hand, did the support for SNS drop?

Let me discuss these in reverse order.  For SNS the results were actually highly consistent with the polls (with the occasional exception of FOCUS and the constant exception of the consistently-errant surveys by Median) which showed a long-term drop toward 5%, motivated I suspect (though I can’t at the moment prove) by the party’s ever-lengthening list of scandal and outrageous remarks.  The question on many minds, I’m sure, is “Why didn’t the affairs surrounding the new Hungarian government” cause its preferences to increase.  I cannot be sure without looking into post-election polling numbers, but the answer may be that these simply did not resonate with voters in the way that they resonated in the media.  Corruption appears to be more tangible and distasteful than statements from Budapest which do not have any clear personal impact for most voters.  The other question is “where did the SNS voters go?”  Here the evidence suggests that as many as 1% of them went to Nase Slovensko, with its even more radical solutions, and that nearly all of the rest either left the electorate or went to Smer.  Voter choice is always relative, and for a voter with a national orientation who nevertheless dislikes Slota or the corruption of his party, Smer is the next best national alternative and less touched by scandal.  This, however, raises the further question of “where did the Smer voters go?”

From one perspective, Smer’s results are “high” only by the standards of the final week of polls, which we now know to be (for reasons unknown) in sharp error.  Leaving aside the final week drop (and the polls of the final week proved in both 2006 and 2010 to make worse predictions than those from a month before) Smer lost more than 10 percentage points from its peaks during 2008.  A variety of experts including Martin Slosiarik at Smer argued at the time that many of these were “soft” supporters who chose Smer as the default option which was in government during a period of economic growth.  The work of Andrew Roberts and others shows that incumbent parties in Central Europe do tend to lose support during periods of economic decline, and many of those softer supporters appear have been affected by increasing unemployment and scandals involving those close to Smer.  Where did they go?  Some simply did not vote, others voted for SDL or KSS (though together not many more than voted for KSS in 2006) and a significant number of the remainder appear (at least in pre-election surveys) to have opted for SaS.  From a purely left-right ideological standpoint this shift seems unlikely, but if voter choice is relative and often non-ideological, then those who simply seek economic opportunity (without having a firm idea of how it should be brought about) and/or seek a cleaner alternative (the role that Smer itself played in 2002 and 2006) then SaS may be a reasonable choice.  (More on this using pre-election data here: http://www.pozorblog.com/2010/06/who-from-whom-slovakia-electoral-shift-roundup/).

Having said all that, I need to point out that Smer not only got the most votes in the election by more than a 2:1 margin but also that it increased its vote share by a sold 5 percentage points over the previous election (and its number of votes by an even greater margin) despite the fact that new parties in government usually lose significant support or collapse completely.  From an electoral perspective, Smer clearly did many things right: it stayed away from the worst of the corruption scandals (perhaps by putting the “ministries of corruption” in the hands of its coalition partners), it took a relatively popular national position without taking it to extremes (though it started to move in that direction in the last year), and it continued to criticize the wealthy and promote redistributive policies (even if many of those were more symbolic than economic).  My question is what happens to Smer now.  It has always been a party that seeks opponents and it will find this easier to do from opposition, but it will face at least three big challenges:

• First, it managed its increased result with a significant inflow of voters from SNS and HZDS and with rhetoric designed to attract those voters.  It would not surprise me if the Smer electorate has not now moved significantly told the older and rural side of Slovakia’s demographic spectrum, ending up where HZDS did in about 1994 or 1996.  This will shape the party’s appeals, as will the weakness of SNS, and it would not surprise me to see Smer move even more fully into the ideological space formerly occupied by HZDS in the late 1990′s of “the (not-as-radical-as-Slota) defender of Slovakia’s national sovereignty.”
• Second, Smer will be going from government to opposition with a very large parliamentary deputation.  It has been in opposition before and stayed very disciplined, but not with such a large group.  It has had a large delegation before but not without the carrots (and sticks) of parliamentary and government offices.  It will be interesting to see whether Smer can avoid splintering if some deputies, perhaps with savvy media advice and outside financial support, see an opportunity for doing better on their own, particularly if the Smer itself inclines more toward the national appeals.
• Third, Smer’s own leader may be torn about what to do in 2014.  Since 2008 I have heard persistent rumors that Fico would rather be president than prime minister.  Because I work from the presumption that leaders would rather have more power than less, I have always discounted these rumors as either wishful thinking (by Fico’s opponents) or misdirection (by Fico’s supporters) but they have come up so often from so many sides that I sometimes have to wonder.  If it is true that Fico would rather be president than prime minister, he will have his best chance to do so in 2014.  Indeed it is hard to imagine a candidate who could come close to beating him in a one-on-one race, and he will have the advantage of running from opposition without the burden of responsibility for government policies.  My suspicion still is that with those advantages any politician I know would still rather be prime minister than president, but if Fico does opt for the presidency (or even lets his mind wander in that direction), then Smer will need to deal with the tensions among a rather large and diverse group of second-tier politicians–Kalinak, Madaric, Caplovic, Paska, Pociatek, Benova and a few others–who may be looking to step into Fico’s shoes and who may not like it much when one of the others takes the spot.

How do you evaluate the results of the two parties representing the Hungarian minority? Were the results surprising for you?

This one has always been opaque to me, in part because I cannot not read the Hungarian press and because what I read in the Slovak press does not even allow me to pretend to know what is going on in the Hungarian community, and there were no precedents that would have allowed me to build a rough model based on past election data.  In principle I found it highly unlikely that the Hungarian parties would maintain the 55:45 split they needed to both stay viable, but the polls pointed consistently at their near equality.  If we discount OKS, whose preference votes accounted for almost 10% of all Most-Hid’s preference votes, the actual ratio of Most-Hid to MKP-SMK among Hungarians was probably about 7.2 to 4.4, so the final ratio was just a bit above 60:40.  It may be that Bugar’s geniality and moderation were more of an electoral motivator than Csaky’s better organizaiton, but for a better understanding of why 20% more Hungarians favored Bugar’s new party than Csaky’s more established one, I look forward to a thorough and demographically-grounded analysis from Hungarian-speaking experts.

I’m off to bed unless the final precincts come in in the next five minutes.  Tomorrow look for a quick roundup that will include a look at party system size and volatility and a look at the relative success or failure of particular pollsters and other methods (hint: say yes to Polis, MVK, bookies and my own “two months out” model; say no to Median and AVVM.)

And for the first time Slovakia will have a female prime minister.  After years of male-dominance unusual even for central Europe, this would be a welcome change.