PNG PM re-elected

Michael Somare, the first prime minister of Papua New Guinea to serve a full parliamentary term, today begins a second term. His party won just 27 of the 109 seats in recent parliamentary elections, but with a coalition of other parties and independents, has now been re-elected to the post by the new parliament.

That political engineering, meant to generate more governmental stability in perhaps the world’s most fragmented democracy, certainly seems to have worked!

PNG, with about six million people and continuous democracy since independence from Australia in the 1970s (and before independence, actually) is one of the overlooked cases of developing world electoral politics. It is also one of the world’s great Law-breakers–Duverger’s Law, that is.

0 thoughts on “PNG PM re-elected

  1. The vote was huge, 86 of 109 MPs voted for (ahem) Grand Chief Sir Michael Somare.

    The PNG constitution goes to great lengths to ensure that parliament is supreme. Actually, the constitution goes to great length on almost every subject, it must be one of the longest in existence.

    The PM is designated by parliament before being appointed by the governor-general. A vote of no confidence must name a new PM and passage of that vote makes the designee Acting PM without any action by the governor-general. The opposition is bringing court action to have the designation by parliament declared invalid.

  2. > “It is also one of the world’s great Law-breakers–Duverger’s Law, that is.”

    Duverger’s Law, like Ptolemaic astronomy or Newtonian physics, still holds, but within a limited field. PNG is outside that field.

    Plurality voting (apart from Approval), as a barrier to new parties and candidates, is like a glass window. Once it’s smashed, it’s gone. If there are only two major contenders, polling an average of 45% to 51% each, and you aspire to to overtake them both, you need to jump from a base of zero to about 35% or 40% to have any chance.

    OTOH, if there are are already a dozen major contenders (as in PNG or Northern ireland), polling an average of 7% to 15% each, and you aspire to overtake them all, you need to jump from a base of zero to about 10% or 15% to have a chance. You can get that far if you just round up all the protest votes.

    Whereas with STV, List-PR or Approval, given a set number of seats, the threshold is only very slightly varied by the number of parties or candidates standing. With STV in one seat (IROV, AV), having ten rivals instead of two means you could win with fewer first preferences, but you still need 50% support in the final count.

    The systems, which are much less affected (to a degree below the “radar” for tactical exploitation ex ante) by vote-splitting, are more like giant foam mattress than a glass window; they may “bend”, but don’t “shatter”.

  3. Ie, Duverger’s Law Mark 1.0 was, boiled down, “Electoral systems with plurality voting and/or single-member districts [1] encourage two-party systems.”

    Duverger’s Law Mark 2.0 should read more like:

    “Electoral systems that set high (a) thresholds for avoiding immediate elimination and/or (b) quotas for ultimately winning a seat, [2] encourage two-party systems.”

    [1] Plurality and single-seaters usually coincide, but but not always: c/f US Electoral College and the French and Australian lower houses.

    [2] For plurality, elimination threshold and exclusion quota are the same thing (50% if two candidates, 33.4% if three evenly-matched candidates, and so forth). For STV, the threshold drops with more candidates, but the [Droop] quota stays constant. For List-PR, OTOH, the threshold (if not equal to the quota) is higher: eg, “we have 600 Bundestag seats, but you don’t get any unless you poll 5%; 0.16667% is not enough”.

  4. Hmm. Actually, there’s not just the “election quota” and the “exclusion threshold”, there’s also:

    (a) “absolute [or a priori] quota” = if you poll this percentage of the votes, you’re guaranteed a seat [1]. The percentage is calculated as [100% x NUM VOTES PER VOTER], divided by [NUM SEATS + NUM VOTES PER VOTER]. This gives us the Droop quota for STV, SNTV and List-PR; it gives us 50% plus 1 for plurality, block vote and (if you can persuade your supporters to “bullet” or “plump” for you) Approval; and it gives us a fluctuating percentage for Limited Vote (eg, 3 votes for 8 seats = 300%/ 11 = 27.27273%).

    (b) the “empirical [or a posteriori] quota” = 1 vote more than the highest-runner-up.

    So, eg, if talking about a candidate’s or party’s chances under STV, we’d say “Well, with 7 seats and three established rival parties, you’ve got no chance at all if you poll less than about 5% of first preferences. If you get 12.501% of first-preferences, you’re home and hosed, but in practice experience shows you need about 8-9& of first preferences to have a chance of winning a seat

    [1] To be precise, you’re allocated NUM VOTES PER VOTER] (or all remaining seats, whichever is less) every time your party has the highest remainder, after subtracting an “absolute quota” in return. So, with 3 votes for 8 seats, parties-wise it’s as if the seats were allocated in batches of 3, 3, and then 2, subtracting 27.27273% from each party’s total every time it was next in line.

    Okay, I think I’ve mystified everyone, including myself… Someone please get us back to the topic of the PNG election with an anecdote or two about Highlands campaigning!

  5. From the Australian Labor Party’s observer mission report:

    Many interlocutors stated, however, that LPV had taken the “heat” out of the election. The capacity of the system to lessen electoral violence because candidates must broaden their vote-share needs some level of systematic scrutiny – the strategies of parties and candidates are too varied to make any concrete claims about systemic behavioural change. Compelling justifications for a down-turn in electoral violence need also to take into account a raft of contingencies including significant police presence, police reform and so on. While an appealing concept, the ability of the LPV system to deliver more peaceful elections needs careful scrutiny.

  6. Funny. I’m not that old, but I can recall the days when the Aust Labor Party (in Queensland) was claiming that the rivers would flow with honey if that State replaced the Bjelke-mander with equal single-member electorates (maximum 10% variation) and optional-preferential voting.

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 )

Google photo

You are commenting using your Google 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.