Malta, election 2013

There is a general election in Malta on 9 March. As there is occasional interest in STV systems around here, I thought I would offer as a public good a proper place to discuss…

21 thoughts on “Malta, election 2013

  1. Polls have closed (10 PM local time close, one of the latest in the world). I have not seen final turnout numbers, but the 2 PM numbers were similar to turnout in 2003, where 96% of registered voters cast their ballots–without mandatory voting.

    Counting will begin at 1000 GMT tomorrow. They start with a sample from each ballot box, meaning parties make projections of seats around an hour later. There are no exit polls that I can find.

    Labour had a 12 point lead at the final polls last weekend. Over a quarter of those surveyed refused to answer how they would vote, but this is apparently the norm. The Malta Today poll in 2008 had the PN winning in a squaker, which was indeed the case. AD-The Greens are polling quite well for them, but they are not expected to win any seats.


  2. The Labour Party have won the biggest victory since 1955, taking 55% and winning by 36,000 votes (around 1400 separated the two major parties a few years ago).

    Labour won the district seats 38-27, including taking traditional PN strongholds in the 12th and 13th district. Under the “strict proportionality in two party parliaments” rule adopted in 2007, the PN will receive two additional seats, to give a 38-29 parliament. At 9 seats, this is the majority in terms of seat numbers in history, and the largest by percentage since 1955.

    The AD slightly increased their total but barring a complete surprise will not be elected.


  3. Sounds like a landslide to me.
    So Malta now has nationwide proportionality? I thought the compensatory STV was just an idea. Were the majority adjustment seats not good enough?


  4. I’m not sure why, but the rule adopted before the 2007 election was that in a two-party parliament, the division of seats would be strictly proportional, with additional seats added on to the 65 district seats. The total number of seats must remain odd.

    In 2008, this meant that the PN received 4 bonus seats. It operated exactly as if the previous rule had been in place.

    In a 3-party (or 4-party, etc.) parliament, if a party with more than half of the first preference votes does not receive a majority of seats, its number of seats is increased to a parliamentary majority. I’m pretty sure that it still is increased to just a one-seat majority and the result is proportional. If, in a multi-party parliament, the majority party already has a majority of seats, or no party has a majority of first preference votes, the district seats stand with no adjustment.

    The largest small party, AD-The Greens wanted strict proportionality with a 2.5% national threshold, using the current STV districts and with top up seats for proportionality. A couple of major party figures proposed national thresholds of 7-10%, but ultimately no threshold was passed and the rule remains a quota in one district (16.7% in the district).


  5. All first counts have been completed.

    Labour won 166,860 votes (54.9%), Nationalist 131,706 (43.3%), AD 5,506 (1.8%, their highest ever at a general election), and others 91 votes. Turnout was 93%.

    Amongst the parliamentary parties, Labour won 55.89% of the vote and the PN 44.11%.

    In the districts where the result is clear, Labour won 3 seats to the PN’s 2 in 3 districts, and won 4 seats to the PN’s 1 in 3 districts. The PN won 3 seats to the PL’s 2 in 3 districts.

    In the 5th district, PL will probably win 4 seats and PN 1. PL had 4.01 quotas and PN 1.91 quotas. PN could gain the 2nd seat due to preference leakage–PL has 13 candidates to only 7 for PN.

    The 8th district’s 5th seat is too close to call. PN won 2.97 quotas and PL won 2.87. The difference between them was 388 votes; AD leader Michael Briguglio won 520 and another AD candidate won 76, so we’ll likely need to wait until Briguglio is excluded to see how preferences flow to determine the 5th seat.

    In the 12th district, PN will probably win 3 seats and PL 2 seats. PN has 3.03 quotas and PL 2.83.

    In the 13th district (Gozo), PN and PL will win 2 seats each and the 5th is too close to call. PN won 3.00 quotas and PL won 2.95. The difference between the parties on the first count was 240; AD had 210 votes. Both parties have small numbers of candidates–5 for PL and 6 for PN, so leakage to exhaustion probably won’t be too bad. A newspaper article cited an “abnormally high level of cross-party voting.” This one will probably be up in the air until the last count.

    If the party ahead at the first count wins all of the too close to call seats, the district seat distribution will be Labour 37, Nationalist 28. If this is the case, I believe that PN will receive 2 adjustment seats, bringing the total to 37-29.


  6. They’re coming from the site Manuel cited.

    Labour have won the 13th district due to Nationalists exhausting their ballots. This is the first time they’ve won a majority of seats in Gozo since 1955.

    It’s also one of the reasons I don’t feel that using 1st count votes to determine proportionality is the fairest–the PN lead on the first count but Labour passed them on the final count.


  7. The final count margin in Gozo was just 9 votes after a recount, with PL taking the final seat with 3,966 votes (213 below quota) to 3,957 for the final PN candidate. 433 votes were exhausted. The PN had a 240 vote lead on the first count.

    The 12th district has also been completed. The Nationalists won 3 seats to the PL’s 2, passing a quota with 4,061 on the final count (the 19th count) to 3,168 for Labour.
    In the 5th district, the PN has 7,697 votes as of the 14th count, with none elected (1.91 quotas). The PL has one elected and 12,231 remaining votes (3.03 quotas), so they should win 4 seats, but the last seat may be with less than a quota. The PL has 4 remaining candidates, and the PN has 3. A Nationalist was excluded at the end of the 15th count, and a Labourite will be excluded after the 16th, leaving 3 more PL candidates and 2 PN candidates for the final 4 seats.

    AD’s Michael Briguglio was excluded after the 11th count in the 8th district with 665 votes. At that point, the Nationalists had 2 already elected and Labour 1 elected. 199 of Briguglio’s votes transfered to Labour (29.9%), 208 to PN (31.3%), and 258 exhausted (38.8%).

    The 13th count is the most recent completed. After this count, the PN had 3,881 votes and the PL had 7,583 with 376 exhausted. A Nationalist with 1792 was excluded and his votes will be transfered. Subtracting one quota of 3,948 from Labour, they will have 3,635 votes, so the PN will most likely win this count, but they have been exhausting at a higher rate than the PL, so it could be close.


  8. The ‘proportionality’ rule explained (full rule, with examples, is in Ch. 354 of the laws of Malta).

    For this election, assuming final breakfown of district seats is 38-27.

    Determine avg votes per seat based on first count votes, distegarding remainders.

    PL 166,860/38=4391
    PN 131,706/27=4878.

    Divide the party with the higher avg’s votes by the lower party’s avg.


    Add this to the other party’s seat total, and round to neareat odd number.

    38+29.99=67.99, so 67 seats. Subtract the PL’s 38 seats, and this gives 29 seats to PN. They elected 27 district MPs, so they get 2 adjustment seats.

    If the seat total were 37-28, the adjustment would result in 67 seats, with the PN having 30.

    How this is ‘proportional’ to the first count is beyond me, but that’s how it works.


  9. > “Chris, could you point us to where these figures are coming from?”

    He digs them up via his own drug-fuelled investigations.

    It’s the school of “Gozo Journalism”…


  10. Haha I had to think for a second about that one. At first I thought it was an uncalled-for slander.

    Labour did take the 4th seat in the 5th district, with all 4 above a quota.

    Labour won the fifth seat in the 8th district–on the final count, PL had one candidate at 3948 (143 above quota) and one declared elected with 3688; the PN candidate had 3680.

    The PN is alleging that a bundle of 50 votes for one of their candidates was placed in the wrong pidgeon hole and that she should not have been excluded, so there may be a complete recount there.

    They have already asked for a recount in the 13th, as they’re alleging that 10 of their votes are unaccounted for.

    9 districts have already completed counting (including those where a recount may take place), and in those, Labour has a 39-26 lead. In the other 4, the specific candidates are up in the air, but the final results are not, and Labour will win those 4 11-9, for a total breakdown of 39-26 (a 13-seat majority). Under the rules prior to 2007, that result would have stood, but under the proportional-but-not-really rule, the final breakdown (assuming the results of the 8th and 13th districts aren’t overturned) will be:

    PL 166,860/39=4278
    PN 131,706/26=5065

    Adjustment: PN 131,706/4278=30.79

    39+30.79=69.79, rounding down to 69 seats. Labour has 39 of those, so the PN gets 30 seats. It has 26 district seats, so there are 4 adjustment seats to the PN.

    Labour would have 56.52% of seats in that case, and the Nationalists 43.48.

    If the Nationalists win both recounts, the breakdown would be:

    PL 166,860/37=4509
    PN 131,706/28=4703

    Adjustment: 131,706/4509=29.20

    37+29.20=66.20, rounding up to 67 seats. Labour takes 37 seats, so the PN gets 30, of which 2 are adjustment seats. Labour would have 55.22% of seats, and the PN 44.78%.

    So much for “proportional representation” when the adjustment is completely different depending on who wins certain district seats.


  11. I would bank on STV operating more consistently in larger constituencies, so the law of large numbers kicks in. Having said that, it does seem that large population or area) for a polity correlates to non-adoption of STV so presumably there is some deterrent factor. ACT and Tasmania are Australia’s two smallest jurisdictions in area and two of the three smallest in population. (Although NSW did use STV briefly in the 1920s-30s.) Malta is Europe’s micro-state and Ireland is one of the smallest in population.


  12. Malta has a considerably greater population than Andorra, Liechtenstein, Monaco, San Marino or Vatican City and is really in the same class as Luxembourg (419 000 v 517 000). The Duchy of Grand Fenwick does not publish census data for some reason.


  13. Alan, I’m as loud a shill for STV as anyone, but saying “STV will work fine in California or Ontario because it works fine in Malta, which has a larger population than, uh, Vatican City” is not a vote-winning soundbite.


  14. To clarify, I feel the position of STV advocates at this point in history is not unlike that of advocates of republican democracy in 1500.”Yeah, right… Switzerland? Venice? San Marino? Not Russia, France or Germany?”


  15. Nor is it a vote-losing slogan to say that STV works in NSW and Australia which have populations comparable to Ontario and California. Nor that New York City had a population comparable to many states and countries in the time their city council was elected by STV.

    The proposition was that STV is associated with small populations. Not only is Malta by no means the smallest European state in terms of population, it also enjoys no special immunity from the rule post hoc non propter hoc.


  16. I have a proposal to make. I was thinking about the STV proposal that was considered by many in Malta. You have multi-member districts, where you allocate many seats by STV and the rest by modified d’Hondt.


  17. Of course, I think what would be simpler, and more reflective of Malta’s STV heritage, would be to use a mixed member-STV system: voters would continue to vote in 5-member local districts and also elect some nationally by party STV.

    You’d cast a national ballot, and determine a national Droop quota, then subtract one quota from the vote total for each district seat won. Because of the small number of parties, preference distribution at the national level wouldn’t be necessary in most elections, but in a very close election like 2008, you’d either see AD elected to the balance of power or else their preferences determining which party gets a majority, rather than relying on the plurality district-level first preferences alone.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.