Time from last votes cast to official results?

[It took me a few days to notice that the subject line did not say what I meant it to say. Fixed now.]

I wonder if my readers can enlighten me on what the norm is for the time between the close of polls and the release of full preliminary official results. I am particularly curious about developing countries and new democracies, and especially those with large and difficult territorial expanse.

I ask because the long gap in Indonesia–apparently it will be about two weeks before official results are announced–is surely a contributing factor in the bubbling crisis over dual claims of victory in the 9 July presidential election. The “quick counts” (samplings of polling-place counts) point to a victory by Joko Widodo (Jokowi)–or at least the credible ones do.

Even for a country as vast as Indonesia, two weeks seems like an unnecessarily long time for a result in the current age, especially when there is only one office on the ballot. I can understand the long delay in a place of ongoing conflict and severe underdevelopment of infrastructure, such as Afghanistan, which had its presidential runoff on 14 June but has no full results yet. And I further understand that systems of paper ballots take longer than electronic voting, such as India and Brazil. But Colombia produces same-evening results on a paper-ballot system with rugged terrain (even if mostly mainland, unlike Indonesia) and with significant conflict zones. It seems Indonesia could do better–and, to mitigate crises over conflicting claims–needs to do better.

But what is the norm?

16 thoughts on “Time from last votes cast to official results?

  1. I’d think the common practice is as soon as possible. The UN ‘expert constitution’ that we see reflected in the constitutions of South Africa and Kenya requires that elections be declared at the earliest possible time. The Kenyan version also requires that each polling-booth be counted immediately and the local results be displayed immediately.

    According to IFES, Indonesia counts the polling booths in daylight and publicly. The delays happen as the booth results are collated upwards through 5 levels with provision for court proceedings to resolve disputes. I would guess it simply didn’t occur to anyone that a losing candidate would take advantage of delays in the system to try and seize the presidency.

    Obviously, it should be the norm that an election is declared as soon as possible, even if that means dispute resolution happens after the result, not before. Separate issue, but it should perhaps also be the norm to minimise the delay between election and assumption of office.

  2. Agreed, it should be as soon as possible, and the entire system should be designed to make quite soon possible.

    I do agree that the delay between election and assumption of office also should be minimal, although with enough time allowed for reasonable challenges to be processed. I am not sure what that is, but there would be such thing as too quick an assumption of office.

  3. Couldn’t Indonesia have a rolling election process like India and after all the stages are finished, then a few days later, all the voters are counted and the result is known instantly?

    Don’t countries that use the STV system (Ireland, Australia, and Malta) have delays in counting of votes because once the votes are counted, then votes have to be transferred to attain the final result?

    • While it does take a few days to count the votes, I believe a large part of the delay in Australia is because postal votes have two weeks to arrive after polls close.

      Even then, most seats are known the night of the election.

      • What about close elections such as the 2010, which produced a hung parliament?

    • Rob,

      As Mark says, the AEC has to allow just under a fortnight for postal votes to come in. The further delay in the Senate is partly due to the fact that all the below-the-line votes (up to 110 numbers per ballot paper in 2013) have to be manually entered into a computer twice. When all that is done, the actual counting, including the transfer of preferences, is very quick. I expect it takes the computer minutes, if not seconds.

      It was actually quite funny reading some of the guff written in the UK about Australia’s voting system when the alternative vote referendum was on.

      We pretty much know the House result on the night, with only around half a dozen seats in doubt. The Senate is less clear because the order of transfer of preferences can come to down to very few votes, meaning that the below-the-line voters, few though they be, can influence the result.

      • The big problem, for which the major parties are largely responsibly, is that the framework of the Commonwealth Electoral Act 1918 is now almost a century old. The postal voting rules made sense in the horse-and-buggy age for which they were enacted. Unfortunately the major parties believe that the postal voting rules give them an advantage over the minor parties so they are not deeply interested in either fairness or electoral certainty. There is no obvious reason why postal voting can’t be deal without the organisational involvement of political parties except that it advantages parties with more boots on the ground. It also leads to a throughly nasty business of party volunteers trawling through hospitals and agecare homes seeking whom they can devour.

        The current parliament will probably further amend the electoral act to deal with the GVT/microparties disaster at the last election. On the other hand I’m told that the cabinet are a lot less enthusiastic for the reforms agreed by the joint standing committee on electoral matters than the LNP members on that committee. What’s needed is a completely new act.

    • It might make sense for Indonesia to use a phased election process like India (and Egypt), given the size of the country and the way it is spread across multiple islands with often very rugged terrain. However, this alone would not solve the problem of delays in getting a preliminary official result. It should be noted that, while different parts of India vote on different days over a long period of time, no votes are counted till all have been cast. (This distinction is what I intended to get at with the subject, although it took me till today to noticed that I had botched the original subject.)

      India uses electronic voting machines, and they are only switched on and their data downloaded after the last votes have been cast. The Indian phased process is a solution to the problem of administration of election day(s), and specifically to providing security, not to the challenge of getting votes counted quickly. And the phasing of election dates predates the use of the machines.

  4. Australia’s longer counting time is a joint result of (1) more generous postal vote deadlines, and (2) preferential counting – especially PR-STV for the multi-seat elections. With IRO-AV, electoral officials could conclusively exclude the lowest candidate if and once (say) only 100 of the postal ballots sent out haven’t arrived back yet, as long as more than 100 votes separate that candidate from the next one higher. With PR-STV, though, you can’t distribute surpluses until you know exactly how many valid votes were cast; you need to know the quota and the transfer value to several decimal places.

    You might be able to eliminate a few of the minnows if no candidate has a quota on first preferences, but that is extremely rare – if not non-existent – in Australian PR-STV elections, whether the closed-list (mainland upper houses) or the open-list (Hare-Clark) versions.

    On this topic, Dr Kevin Bonham http://tinyurl.com/kv7jbuk has noted that Hare-Clark (at least the “last parcel Gregory [fractional] transfer” version used in Tasmania) is less vulnerable to the entire result being cast into doubt by a few compromised ballots than are other STV variants – specifically the Senate’s Uniform Inclusive Gregory method (hence the 2013-14 West Australian re-poll) but also Meek and Weighted Inclusive Gregory.

    I’m a bit of a lone heretic among Australian STV shills for supporting “first-parcel Gregory [fractional] transfer” – in other words, you start by examining the elected candidate’s first-preference ballots, then (if the number of un-exhausted first preferences is less than the surplus), those he or she received on the second count, and so forth. This variation…

    (1) would be less arbitrary, ie, fixed at the start of the count and not fluctuating according to later order of elimination.

    (2) would minimise the incentive not to waste one’s vote on someone likely to be elected earlier in the count. It would not give a positive incentive to vote for an early winner, as (non-weighted) Inclusive Gregory does (with its prospect that one’s ballot might increase in transfer value to more than its initial 1.000 votes), but at least it reduces a disincentive. Your vote will probably reduce (or at most, remain unchanged) in value if it’s selected for transfer, but it has a much higher chance of being selected for transfer.

    (3) inasmuch as elected candidates have more votes than eliminated candidates, basing surplus selection on the former is less likely to ride on tiny numbers of votes separating the 22nd-lowest from the 23rd-lowest candidate.

    • I want to know which of the numerous STV shills on this blog bribed you to engage in this explicit act of threadjacking.

    • That’s unfortunate. I guess more micro parties will win seats on the basis of the preferences whisperer.

      • It’s not unfortunate, it is thoroughly reprehensible from a government that made hay during the minority governments by denouncing ‘dodgy deals’ with the cross bench.

        It is impossible to imagine a dodgier (if unstated) deal than guaranteeing the microparties continuing access to preference harvesting in return for their votes in the current parliament.

      • Rob,

        It’s not unfortunate at all. The whole argument was always irrational, based on hysteria that a few outsiders managed to do what hundreds of the number two and three candidates of the major parties have been doing without controversy for the 65 year since PR was brought in as the Senate voting system. I have always said that the Coalition might eventually wake up, realise that the party to benefit from the “reforms” would be the Greens and decide it really wanted an alternative to be able to hold the balance of power in the Senate.

        I won’t hijack this thread into a discussion of Senate voting “reform”, but invite those who want to discuss it to go to Western Australia Senate re-run and the Joint Committee on Electoral Matters.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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.