Projection: SNP majority of 31, Willie Rennie to lose seat

mar 1 tns

SNP: 80
Labour: 28
Conservative: 13
Lib Dem: 5
Green: 3

TNS 1 March 2014

This week’s TNS poll brings great cause for cheer for the SNP, reaching new heights in both constituency and regional voting intention – polling at 60% and 55% respectively. Labour are on 21% by both measures, which at least brings the comfort that they don’t appear to be declining any further (small comforts, eh?)

On the latest figures, I’m projecting SNP on 80 seats, winning a thumping majority of 31. 28 seats for Labour is at the higher end of recent expectations of the party’s performance. The Greens gain a second seat in the Lothians, and the Lib Dems remain static. Perhaps, most surprisingly, the much-touted Conservative surge is still yet to materialise – with a projected 13 seats, a 2 seat decline on 2011.

Once again, the South of Scotland provides some of the most interesting data, with Tory support there plummeting. Perhaps by broadening the Tories appeal to newer voters, Ruth Davidson is alienating some of the more traditional support in Tory seats.

The Tories are projected to lose all of their seats in the South of Scotland, but gain Eastwood from Labour. The Lib Dems retain Shetland, but lose Orkney, and are compensated with a seat on the Highlands and Islands list. On the basis of the TNS poll, it’s a constituency wipeout for Labour.

On a regional basis, Labour is projected to win 28 seats. Central Scotland is projected to be Labour’s strongest performance, winning 5 seats, with 4 in Glasgow and West. While the Liberal Democrats remain static on 5, their leader Willie Rennie is forecast to lose his seat in Mid-Scotland and Fife. the Lib Dems are projected to pick up regional seats in Glasgow, Central, and Highlands and Islands – meaning returns for former MSPs Robert Brown and Jamie Stone. Perhaps most astonishingly, the SNP wins FOUR regional list seats in the North East, in addition to sweeping all 10 constituencies. 

The resurgence of smaller parties projected by the previous Survation poll isn’t borne out by TNS, with the Greens forecast to pick up only one more seat in the Lothians.

Sturgeon’s astonishingly risky EU intervention

SturgeonThe EU referendum was always going to be a tightrope for Nicola Sturgeon. The risks are manifold: saying something that could ultimately be used against her in future Scottish independence referendum; drawing attention to some of the more unpleasant elements of the independence cause; and appearing to stand shoulder-to-shoulder with the Prime Minister, having previously denounced (and demolished) Labour for doing the same. I always therefore assumed that come the independence referendum Sturgeon, being more risk-averse than her predecessor, would restate her position that she favours British membership of the EU and then stay out of the debate. However, on Monday she delivered a most astonishing speech on Britain’s membership of the European Union – exposing herself to risks on all three fronts identified above, and more.

She acknowledged, though did not dismiss, the inconsistencies in arguing for a “pooling of sovereignty” with other EU Member States while not applying that same logic mutatis mutandis to the United Kingdom. The submission that “there is nothing at all contradictory about independent nations recognising their interdependence” is so packed with contradiction that only a true-believer could fail to spot it.

In the independence referendum the SNP hoped that we would suspend our better judgement and vote with our hearts, rather than our heads – and dismissed the pleas of those who urged us that we do opposite as cynical and negative. The irony of now making the instrumental case for a political union while dismissing the more emotive, nationalistic case for Brexit is clearly lost on her.

Sturgeon made the case for the EU imposing certain certain social standards on the UK (purportedly) against its will:

“In 2013 the UK only increased the minimum entitlement to parental leave as a direct result of European directives. There are other cases – for example minimum annual leave, and conditions for agency workers – where the UK complies with the European minimum and no more. Which begs the question without European regulations would minimum standards be meet the regulation at all?” [sic]

In other words, it’s OK for the European Union to impose laws on the UK that the UK has not voted for, but it’s not OK for the United Kingdom to impose laws on Scotland that Scotland has not voted for. I wrote previously about how the left was generally fine with surrendering control over the levers of power provided the levers are nonetheless pulled in a manner of their pleasing (such as the internationalisation of human rights). However, pulling the levers of power is the core purpose of the SNP and the independence movement. The SNP didn’t give up its pursuit of independence when Gordon Brown was Prime Minister even though Brown clearly enjoyed overwhelming support in Scotland. The fact that the levers of power were being pulled in a manner Scotland chose mattered not one jot to the SNP – what mattered most was who was pulling the levers of power, not how they are pulled – which is where the contradiction in Sturgeon’s position lies.

She continues:

“In fact, when you consider some of the UK Government’s other policies – for example its attempt to further weaken trade union rights – we should be thankful that the European Union sets some basic social standards.”

The 2015 General Election was remarkable for a number of reasons, though one result seldom commented upon is the fact that right-leaning parties won a majority of the vote for the first time since the war (though it was close in the 1950s). For Sturgeon to claim that we Britons should be thankful to the EU for imposing policies upon us that the people of the UK voted against at the last election, while simultaneously crying foul every time the UK pursues policies which “the people of Scotland did not vote for” is bare-faced hypocrisy, and impossible to credibly sustain. The very nature of a political community is one in which individuals and groups subjugate their own desires to the will of the community. The fact that Sturgeon is apparently OK with the principle of laws being imposed against a nation’s democratic will provided those laws come from Brussels and not London, is a conflict that cannot be resolved without exposing some fairly unpleasant truths about some aspects of Scottish nationalism.

Sturgeon further claimed that

“[t]he European Union is good for the prosperity and wellbeing of individuals, families and communities across our country.”

This is a questionable assertion, but whatever the truth of it in the UK, it is most certainly not the case in Greece, Ireland, Portugal, Spain, or Italy – where the EU has imposed swingeing austerity on these nations, usually against their will, in order to protect (primarily the financial industries of) Germany and France from contagion. How can a party that claims to be “anti-austerity” be so enthusiastically supportive of political institutions that have imposed the harshest austerity Europe has seen since the war? Sturgeon’s narrow, self-interested, view of the beneficiaries of the European Union doesn’t square with her support for a social Europe. I would also ask this question of Jeremy Corbyn and John McDonnell, though their support for the EU is, at least, considerably more tepid. This is a contradiction that Sturgeon conveniently glossed over in her speech.

Sturgeon tail-ended her extraordinary foray into the EU referendum debate with an interview with Channel 4 News. Sturgeon was asked what her supporters might think about her fighting on the same side of a referendum as the Prime Minister. She replied

“[i]f he can appeal to those who agree with him, I’ll appeal to those more inclined to my progressive social democratic view of the European Union – if that maximises the vote to stay in, all the better!”

Substitute “United Kingdom” for “European Union” and that line could as easily have been uttered by Jim Murphy or Johann Lamont.

None of this is explosive, though it certainly provides the SNP’s critics (for what they’re worth) with some ammunition. What’s most remarkable is that such a risky speech is entirely out of character for cautious Nicola Sturgeon. Plenty of other political leaders who have won landslide victories became over-confident. Ultimately, it was their over-confidence that was their undoing.

Machine counting of STV elections: faster and fairer


As we face the prospect of a fourth day counting votes in the 2016 General Election, it is time for Ireland to consider following Scotland’s lead and implement machine counting of single transferrable vote (STV) elections. First and foremost, I am not talking about revisiting the failed electronic voting experiment. I am quite sceptical about such systems, as they rely on the integrity of the technology with respect to not only the count, but the very process of voting itself. What I am talking about is the counting process. Machine counting of STV elections has taken place twice in Scotland, most recently using systems supplied by a Derry-based company. The process is significantly faster, with counts taking a few hours rather than a few days. It is also, perhaps crucially, significantly fairer.

How does it work in Scotland?

Votes are cast in exactly the same way as in Ireland at present. You simply mark your ballot paper and place it in a ballot box in the same manner you always would records are kept at the polling station of how many people have voted, and should correspond to the number of ballot papers in the box once it is sealed. So from the outset, there is a paper-based poll that can always be hand counted in the event of any systematic failure.

The poll is verified by hand

The poll is verified by hand

Once the count commences, the ballot boxes are opened out onto tables and the poll is verified (checked to ensure that the number of ballot papers in each box corresponds with the number recorded by the polling station) by hand. This process involves counting the ballot papers into bundles of 100.

It is after this process that the machines kick in. The bundles of 100 are placed on top of a scanner and fed in one at a time. A computer then uses optical character recognition software to read the preferences indicated on each ballot. The counting agent operating the computer has to manually verify that the computer has correctly identified all of the preferences indicated on the ballot, and correct any obvious errors in the recognition. Any ballots that are in any way questionable or potentially invalid are referred for adjudication, just as would happen in a hand count. The process typically takes around 3-4 seconds per ballot – so longer than a hand sort – but, crucially, you only have to do it once. The whole process can be overseen by the candidates and their teams. The computers have two screens – one facing the operator and one facing outwards. You can see each ballot as it is fed in and verify that what was on the paper is what’s on the screen. Candidates’ agents still tally votes just as they would a hand count.

Count staff feed ballots into the computer, while candidates and agents oversee the process

Count staff feed ballots into the computer and verify the recognition of preferences. Candidates and agents can oversee the process on computer screens.

However, once all of the ballot papers have been fed in, checked, and disputed ballot papers adjudicated upon, the computer can then run through the whole election count instantly. In my home local authority, STV election counts take the same length of time it takes to conduct a simple first past the post count for the same electoral area, with the same size of count staff, in general elections (approx. 3-4 hours).

fair distribution of surpluses

Aside from the expediency of machine counting, the system is also much, much fairer. A difficult problem with STV is when an elected candidate has a surplus to redistribute, how do you decide which of their votes are the surplus ones to be redistributed? In Ireland, this is done at random. This is obviously grossly unfair.

Consider the example of Dublin South-West. John Lahart was elected at the 11th count with a surplus of 190 votes out of 11,402. If a newspaper were to print a poll with a sample size of 190 it would rightly be ridiculed for being so small as to be in no way an accurate reflection of the sentiments of the electorate. If such a small sample is unacceptable for hypothetical elections in opinion polls, it certainly unacceptable when real votes are concerned.

An example from a local election in Scotland, illustrating the fractional allocation of a candidate's preferences.

An example from a local election in Scotland, illustrating the fractional allocation of a candidate’s preferences.

The much fairer way is to count up ALL of the next preferences of that candidate, and then prorate them according to the size of the surplus. The result is fractions of votes that perfectly reflect the second preferences of the whole of that candidate’s vote. You can do this by hand, but in the above example it would mean counting over 11,000 votes just to allocate 190. It is hugely time consuming. A computer, however, can do it instantly.

So not only is machine counting quicker, it also makes it much easier to allocate surpluses more accurately.

Can it be trusted?

It has been put to me that such a system requires a lot of trust – in the software, the hardware, and the supplier – which is true. Just as the existing system requires a lot of trust in hundreds, if not thousands, of polling and counting agents. We have no reason to believe that counting agents would ever try to manipulate a count, nor should we have any doubt that the makers of the software or hardware would. There is always the possibility of error, which is evident from the recounts presently ongoing in the General Election. But while human error is random and unpredictable, making it much more difficult to spot, systematic error is, by contrast, comparatively easy to identify. Parties will still have been tallying votes at the verification, and overseeing the scanning process, so discrepancies in the system should be easy to spot. The actual code required to allocate votes once they have been fed into the system is incredibly simple, which a freshman computer science student should be able to write. It should be similarly simple to independently verify the veracity of such software.

Ireland is not exceptional. If machine counting of paper ballots works well elsewhere, there is no reason why it wouldn’t also work in Ireland. It is most certainly not prohibitively costly (in particular when you consider how much counting votes for four days must cost). Machine counting is many times faster than hand counting, and more accurate. Ireland should put the failed electronic voting experiment behind it, and follow Scotland’s example in implementing machine counting of STV elections.