CityCyclingEdinburgh was launched on the 27th of October 2009 as "an experiment".
IT’S TRUE!
CCE is 15years old!
Well done to ALL posters
It soon became useful and entertaining. There are regular posters, people who add useful info occasionally and plenty more who drop by to watch. That's fine. If you want to add news/comments it's easy to register and become a member.
RULES No personal insults. No swearing.
closedCan't remember. Mr Salmond left the 3rd option on the table but Call Me Dave and the other Noers didn't want it because they thought it was Mr Salmond's second prize.
It could so easily have been No's first prize but they were too smart for that.
Voyager had some good episodes, mostly those featuring the Doctor but DS9 had In the Pale Moonlight, Far Beyond the Stars, The Visitor, the nine episode arc and anything with Garak as a main character. They even did The Searchers and Magnificent Seven.
It seemed like the intention was to humiliate the independence campaign with a clear No rather than allow a consolation prize. Of course, it also seems like a good interpretation of wanting it on as indicating weakness, uncertainty, on the Yes side.
But if they wanted a clear No, they should have insisted on a faster referendum. They must look back and think that for the No side Wendy Alexander called it right, with her "bring it on". They should have forced it then rather than allow a slow, relentless campaign on the ground.
As it stands just now, depending on how I mash the data, the trends look like there's a chance that Yes will win. On current trends, they won't win so it'll take a last minute push to lift support a bit but it's within reach.
No surprises there.
It would be a strange sort of independence where many of your laws would be set by one third party (Europe) and your currency by another (UK). Especially if the UK is no longer part of Europe.
And not forgetting the nuclear weapons and monarchy of the UK that you've also had to accept.
Pin,
Have a listen to More or Less on Radio 4 for the European law point. Anyway, I thought the story was that we wouldn't be allowed in.
Similarly we will not be allowed to use the pound. I'm not sure how that would work and why Westminster think it's up to them.
I thought even the MOD had decided that the enclave idea was silly. Would we be allowed to keep the nukes or might that be against anti proliferation agreements? Plenty of sharks to jump on that one.
Brenda will not be around for ever and we could vote to change it when Charlie boy gets the gig.
First draft question might have been Do you agree Scotland should be an independent country?
I agree should have happened sooner now would be fine. Inthink all the arguments have been gone into at some length. And yet some people say they haven't made up their minds. Fallacy on both sides that they can sway these people? They will decide if they go to the polling station, if they have registered.
What might make a difference will be factors on the day (eg poor weather on the day favours Tories in England).
"why Westminster think it's up to them"
This is clear. We can use the pound. (Lots of places use other currencies rather than a national one). BUT, we would the sacrifice a certain element of fiscal autonomy and concomitant sovereignty as a result.
This is why French West Africa often said to have a neo-colonial or dependent relationship with France after independence because their currency linked to Franc.
@Insto, interesting.
The No camp must be getting jittery. Their Loyal press are resorting to character assassinations of Salmond. See Tuesday's Record for instance.
Jittery, maybe but unnecessarily, I think. You have to set aside a lot of data, take all the polls at face value and use the most optimistic projection method to predict a majority in September (which is what I assume Yes Scotland's analysis reported in the Herald as the basis of shifting the point where they'll overtake No from June to September has done).
What no one seems to have done (at least not publicly) is try to calculate the house effects in the polls - the extent to which each company systematically over or under counts each of the Yes, No, Don't know options - and then adjust the polls to take account of it. It's pretty clear that Panelbase overstate Yes support and TNS understate it. The range in the estimates is about 10 points when it should no more than about 3 or 4. Don't know is similarly very variable. On average, 17 percentage points between the lowest (Ipsos MORI) and highest (TNS) measure. But if each survey were measuring the same underlying population, they should be broadly similar. Those differences are methodological and need to be accounted for.
Still, there's no denying that Yes gained about 7 percent since November 2013 but the rate of increase has been too slow and seems to have stalled again.
But all projections assume nothing will change between now and September and there's plenty to go wrong for everyone before then.
Following on from the discussion of certaintly, I've just been looking at a study published by the University of Stirling which, among many other things, looked at risk aversion and likelihood of voting Yes. It says:
"As can be seen, attitudes to the future have no effect on probability of supporting a Yes vote. However, there is a significant and substantial effect of attitudes towards risk. Those with high levels of tolerance of risk have substantially higher levels of support for Independence. For each point move on the 1-10 scale, the probability of supporting Yes increases by 3.6 percentage points."
So, it's no surprise that Yes should seek to make independence positive and risk-free and that No should emphasise the risks and uncertainties inherent in independence. Each is a sensible strategy if risk perception is a major factor in people's choices.
"
What I’m against is politics based on tribalism, be that class, religion or nationalism. And I’m obviously against politics based on dynasty or personalities.
I’m Scottish. I will always want Scotland to beat England at any sport from tiddlywinks on up.
But when it comes to Scottish nationalism in a political sense, I have problems.
I have no idea if Scotland would be better off independent or not. But what I do know is that I want fewer borders in the world not more.
And I don’t want politics that exploit or pander to my more romantic notions of Scotland. I don’t want politics based on a notion of what we think our country once was and may be again.
"
However the further north you go in Scotland, the closer the culture is to Nordic. The further west, it gets more Celtic/Irish.
The same applies to England.
@ morningsider "I was trying to highlight that quoting bond rates, being paid at least two years before Scotland could become independent, by other countries offers no real guide to what a newly independent Scotland may pay."
I disagree with you. I think it's a very good guide to what Scotland will be paying. And I certainly think it's a better guide than the opinions of economists, especially those with a weak track record.
In fact, I find it curious that people would choose the opinions over the comparable realities.
"While Scotland is a small northern European country, it shares little else with Sweden (or other Nordic countries)... we are really far more like England than the Nordic countries due to shared laws, culture, language, currency, institutions and so on."
For one, I disagree with you - taxes aren't anything like as high in Nordic countries as people here seem to think. State intervention is more obvious in some ways - like alcohol - but less in others, defence, policing, prisons.
But more to the point, culture and language have nothing whatsoever to do with the rates you pay on borrowing - these are set by the strength of your economy and the market's assessment of your likelihood of default. Given that Scotland's economy is larger per head the the UK's, and our taxes generated per person are larger than the UK's, and we are more dependent on exports and external income like tourism than the UK, and that we are backed by huge oil reserves, our economy is more like that of the Nordic countries than it is like the UK's.
An even closer comparison would be with Ireland, but without the property bubble and crash they had. Ireland's 10 year bonds are at 0.08% higher than the UK's, and their last 5-year bond issue went LOWER than the UK's.
So can anyone tell me why we'd be paying more than Ireland?
"I think we have already covered the fact that Scottish GDP figures and UK/Scottish public finances are as much a work of interpretation as solid fact."
I think there's a problem with the way you're seeing this.
Correct me if I'm wrong, but it seems as though you're saying that some things are hard facts, but anything short of that is up in the air. The reality of course is that there are varying levels of certainty about the things that aren't facts.
Would childcare be much better in an independent Scotland? There's really not much evidence to say either way, so it's supposition. But the size of Scotland's economy - we may not know it exactly, but the figures are very robust. And figures on public finances may be an approximation, but they will be accurate to a pretty small margin of error.
"Would it really hurt to accept this uncertainty?"
Equally I could say, would it hurt to accept the the evidence for some things that aren't certain is still very strong?
@Instography
""But what no-one should be saying is that the UK is a successful country when it is lagging so pitifully behind its neighbours. Nor should anyone be saying there's an advantage in being in a big country, because that's manifestly disproven."
Did anyone do that?"
All the time! The two most common "positive" reasons given by the No side are, why break up a successful union, and we're better off as part of a bigger country where risk and gains can be shared.
I think the premise behind those assertions is contradicted by evidence.
"It's pretty clear that Panelbase overstate Yes support and TNS understate it."
I don't know how you can say that. One might be spot-on. They both might be overstating, or they could both be understating. While it's possible to check methodology for election polling against previous results, there's no previous result against which to benchmark the polling methodology on the referendum.
The differences between the companies on the indyref are mostly down to factors around weighting SNP support - do you weight against the 2010 general election, or the 2011 Scottish election? Voter recall also becomes an issue, and the shifting loyalties and vote-splitting that now happen in Scotland make it much trickier.
The pollsters have it very hard on this one, it's a minefield for them.
As for polls as a predictor of the final result, I think a good 20, 30% of the electorate won't make up their mind until the final weeks. That's increasingly what happens in election and there's reason to think it will be an even bigger effect in this referendum. People will think "right, I really need to look into this and decide what way I'm voting" and they'll go online or read the papers or talk to people or whatever, and then they'll make their mind up.
It could easily be the case that between 6 and 3 weeks out the polls suddenly shift to 60-40 in favour of one side or the other and stay there. Right now the trends in the polls are interesting, but it's definitely too early to read much into them, other than to say that it's all to play for.
Now with my biased hat on - I think that a majority of Scots have a leaning towards independence but many are worried about whether we can afford it. Most people in this group would say that if the vote were tomorrow they'd vote No - that's the default position. When they find out that Scotland isn't subsidised, and if anything the money flows the other way, they move to Yes.
That's my theory anyway...
"I don't know how you can say that."
Evidently.
"
Right now the trends in the polls are interesting, but it's definitely too early to read much into them, other than to say that it's all to play for.
"
Think that's true!
I wonder how many people saying 'don't know' would/have previously said yes or no - either because they are in transition from one to the other or from certain to uncertain and may go back to their initial intention.
Are there any polls that have a (large sample) panel that they have been questioning for at least a year?
There was a time when it was suggested that whichever side managed to 'prove' that people would be better off would win.
I think that people are increasingly unwilling to believe either side. (Plenty of 'when is a fact not a fact' on this thread!)
I expect there are some people who would be more inclined to vote Yes if they thought that the SNP wouldn't win at Holyrood next time. But that is less predictable than the result of next year's Westminster election.
I think the result of the Euro election - two weeks away - (whatever it is) will concentrate minds.
calmac - I previously posted that:
"You can analyse current "facts", but the results of that analysis will always be an opinion. Some opinions are clearly better founded than others, but that still doesn't make them facts."
I think that's pretty clear. I'm happy to accept some things that aren't clear. For instance, I think an independent Scotland would become a member of the EU - a process would be hammered out to allow this to happen as it is everyone's interest, even though the exact route to membership isn't clear at present. The issue of Scottish borrowing rates is far less clear, as it will be influenced by numerous factors that aren't clear just now. Simply saying that we will pay something like the rate Sweden enjoys just now is too great a leap - there is no firm evidence for this claim.
I based my claims on the difference between Nordic countries and Scotland on personal experience of studying at a Norwegian university. I like to think I have a pretty good handle on this.
I'm not sure why you keep mentioning Ireland, but I would have thought the fact that the rest of the Eurozone (well Germany really) stands behind them has something to do with the rate they pay. Something a newly independent Scotland would not benefit from.
Each poll purports to be a representative sample of the Scottish adult population. Surveys that are undertaken at around the same time should, within the limits of sampling theory, each produce more or less the same result.
The existence of persistent variation in the polls to an extent well beyond the normal range for surveys of their respective sizes indicates a fundamental methodological problem. The mean of the sample estimates should equal the figure that would be found in the population if the whole population were surveyed. The range of the survey estimates should be within the confidence intervals for surveys of that size (they're all around 1,000 sample so the range would be 3-4%).
But here the range is too wide - 10%. So I'm forced to conclude that either all of the surveys are not sampling from same population or that not all of the surveys are properly representing the population from which they are sampled.
If we look at the variation in the survey estimates we can see broad consistency between most of the companies but two outliers. If I exclude the outliers the range of estimates is 4% - exactly what I would expect.
The two outliers are Panelbase and TNS. Their deviation from the consensus of the others is dimensionally and directionally consisent so on that basis I say with confidence (my professional confidence, based on 25 years of this stuff, not to be confused with statistical confidence) that Panelbase overstate Yes and TNS understate it. Since their samples are supposed be from the same population and their deviation is well beyond the expected range, the problem is methodological. I'd call them house effects. No other conclusion is consistent with the theory and the numbers.
Trends. On the contrary. We can read a lot from them. They tell us, for example, in which direction and at what speed opinion is changing. Within a relatively short time frame looking forward, and with reasonable assumptions, they indicate the likely range within which things will move. Locating turning points with events, they give us a sense of the nature and scale of the events that initiate turning points. If we look at how people respond to them then they give us an insight into the mood of the campaigns.
Turning points can be both in the data and in the campaigns. The trends and the availability of data allow us to ponder what happened just before the publication of the story in the Herald where Yes Scotland changed their mind about when they would overtake No. A significant event, I think, based on their analysis of the trends.
@insto very good analysis there of the distribution of poll data. By house results I assume you mean the actual questions the poll asks and / or the select nature of the populAtions sampled. Very interesting to me to note that the skew here is outwith normal distributions of poll results, if I have interpreted this correctly. I am interested in whether an actual narrowing is occurring or if the idea of this is being manipulated. I have a feeling, that is all that we are in the 1992 Westminster election. About 10pm I came back from Edinburgh to Glasgow and came through queen st on my way to maryhill road for the celebration and picked up a paper that said labour largest party in hung parliament. Got into the flat at 124 Maryhill Road brandishing the paper but the comrades said No Gembo, it is worse than that, early indications showing Tory victory and polls worth toffee.
You know where I am going with this but I could be wrong.....
"Trends. On the contrary. We can read a lot from them."
Generally of course. But the issue 'here' is partly about explaining results that don't conform to 'industry norms'.
I don't know if (or how much) the polling for the indieref is different/difficult, but it does seem that there are 'unexplaineds' so perhaps it's harder to divine trends for the September poll.
However, relatively recent trends of (increased) tactical voting suggest that things are changing and (perhaps) it's harder for polling organisers to get the requisite 'balance' without asking more people - which is obviously more expensive for poll commissioners
Tightening? It all depends what you take as your start point. If you look from November 2013 to April 2014, Yes picked up +8 points. That's the maximum since the campaign began. If you look from July 2013 they're only +1 (based on the monthly polling average across all polls).
There's 35 years of data on support for independence in Scotland. It still hasn't reached the high point after the devolution referendum when it was nearly 50%. But it's massively higher than it was in 1979, when it was only 10%.
@insto nearly 50% in the devolution referendum of 1979 do you mean? 51.6 % of a 63.8% turnout falling short of the 40% of the electorate required?
When was it 10%?
No, in 1998, after the 1997 devolution referendum support for independence was 50%. In 1979 it was 10%.
Ta for clarification
1979 support for devolution 51.6% of 64 % turnout
1997 support for devolution on first yes question 75% on 60 % turnout.
Support for independence in 1979 according to opinion polls 10%
support for independence in 1997 according to opinion polls 50%
Conclusions? approx 45% of voters in Scotland wanted devolution in 1997 cf 32% in 1979
Opinion polls for independence show similar movement
This time independence on its own and support in opinion polls has dropped from 1998 high point?
I think we should make the politicians angry and see what happens.
@chdot
In principle polling for the referendum should be no more difficult than for anything else. For precision, you need a relatively large, unbiased sample of the population. That's all.
None of the polling companies use the most rigorous sampling methods - a randomly selected sample of the population. The Scottish Social Attitudes Survey does but that only happens once a year but it's interesting to look at the polls around the time of the SSAS and see what the variation was. (Guess which two polling companies deviated most from the SSAS estimate?)
In the referendum there are three survey methods being used for polling:
* internet panels - almost everyone - Panelbase, ICM, Survation etc.
* telephone surveys using random digit dialling - Ipsos MORI only
* face-to-face in-home - TNS-BMRB only
Each of them is producing samples that broadly match the population, with expected under-representation of young people etc. Nothing particularly alarming and the various weighting schemes they use have pretty limited impact.
Few have a long track record of this kind of thing. Panelbase and Survation are quite new companies and, as far as I know, ICM, although long-established are new to internet panel surveys on Indyref. Their other polling is done by telephone.
Why the variation? We need to speculate since we don't have access to the kind of information that would really let us see what is going on. So, let's speculate. I'm going to assume that when people participate in surveys they generally tell the truth. I know there's something about shy Yes or No supporters but I'm not really buying that.
There's a general problem for all surveys of differential non-participation - that the people who take part are different from people who do take part in some fundamental but non-obvious ways. This is true bias. To illustrate the problem: let's say you want to estimate the prevalence of cycling. One approach might be to put some questions on a standard market research omnibus. The sample composition looks OK but somehow it doesn't seem right. The estimates are lower than some national statistics survey that you compare with. When are market research omnibuses done? Mainly at the weekend. When do lots of people cycle? At the weekend. So even though the survey looks OK, it's estimates are fundamentally biased.
Are there parallels with Indyref. Let's consider the most prevalent method - internet panels. The first question to ask is who the hell joins an internet panel? And how might these people differ attitudinally from the population as a whole?
I'd speculate that people on panels are either (a) in it for the money or (b) highly motivated - interested in political surveys. That makes them automatically substantially different from the population as a whole and opens up the potential for some fundamental bias. It puts significant pressure on the pollster to identify and deal with this problem. This makes the story about Panelbase from September 2013 highly relevant - they had to close their panel amid claims (which they denied) that their panel had been invaded by Yes supporters.
Telephone surveys - they generally do not include mobiles so have a potential problem if people who only have a mobile are attitudinally different from people who have a landline. At a population level the two populations are different but within demographic sub-groups I can't see any logical connection between the type of telephone you use and your views on independence. But there might be. If the survey is introduced as being about a particular topic there's a risk that people differentially refuse. This happened when the Scottish Crime Survey was trialled on telephone. People with no experience of crime tended to refuse so the crime rates in the survey were far too high. So there's a risk but it's predictable and manageable.
Face-to-face surveys. Actually I can't think of a single reason why TNS' estimates should be so wide of the mark. They have buggered about with the question, persisting with some in-house oddity long after the SG's preferred question had been published and after the final question had been settled. Even now, paradoxically, they ask how people will vote in September rather than the standard 'tomorrow' referendum. I suspect that explains much of their DK response but not really why their Yes estimate is so low. Of all the methods, they should be best but are easily the worst.
Sorry, this is a very long answer to a short question. On the one hand it should be straightforward but in a multi-method contentious referendum it becomes phenomenally complex. Me, I cling to the lifebelt of the fundamentals. The question isn't so much which of the samples is biased but how are they biased, where is it coming from and how can we tell by how much.
An acknowledgment from the Herald of the Nationalist intimidation that's going on. You can't blame the media outlet for this one.
http://www.heraldscotland.com/mobile/business/people/haughey-attacks-online-abuse-of-tycoons-who-back-the-union.24161552
This topic has been closed to new replies.
Cycling in Edinburgh Flickr group
Video embedded using Easy Video Embed plugin
posts