Scotland skills 2030: The future of work and the skills system in Scotland

Institute for Public Policy Research (IPPR) Scotland, today:

The world of work in 2030 will be very different to that in 2017. People are more likely to be working longer, and will often have multiple jobs, with multiple employers and in multiple careers. Over 2.5 million adults of working age in Scotland today (nearly 80 per cent) will still be of working age by 2030. At the same time, over 46 per cent of jobs (1.2 million) in Scotland are at high risk of automation. We will therefore need a skills system ready to work with people throughout their careers.

There have been improvements in the Scottish labour market in recent years in terms of pay and productivity, while qualification levels have been steadily improving and are higher than levels in the UK as a whole. However, Scotland continues to have lower rates of in-work progression and lower rates of productivity than the UK as a whole, and pay rates, although increasing, have reduced in real terms and are still behind rates in the UK overall.

Within the skills system, there are gaps and overlaps in provisions, with a clear gap in mid- career provision, which employers are not addressing. While we have the best record within the UK nations for employers investing in training, there is still much to be improved on, with too many employers adopting a low-skill business model.

Chart: Proportion of jobs with high and medium/low potential for automation by region or nation

You can read the full report here.

General Election 2017 projection and prediction, 4 May update

By Jeremy Blackwell

I will have an update on my projection and prediction every week on Thursday until Election Day. I will keep it short and mostly point to the changes since the last update. At the end of each update I will also discuss some specific points like ‘seats to watch’.

How did the polling data evolve?

The complete crosstabs for some of the weekend polls were late showing up. And there were two Tuesday polls to add. But the main trends have not changed significantly since Monday. For better visibility I have hidden the small dots that represent individual polls. The large dots on the left side are the actual 2015 results.


The rolling average barely changed with the new polls in:


Continue reading General Election 2017 projection and prediction, 4 May update

Roundup, Wednesday 3 May 2017

General Election 2017: which way will Scotland go?

By Jeremy Blackwell

You’re lucky this time you won’t have to go through the ‘A few words about me’ and the ‘Hits and misses’ parts. You already got them. So let’s go directly to the heart of the matter.

How does my model work?

Like all election projection models, my model works mostly on ‘Uniform National Swing’ (UNS). With UNS you have all seats moving in exactly the same direction as the national polling average—which never happens in real life.

So I have a second algorithm on top of the first one which I call ‘relative swing’. It factors in not just the usual UNS, but the way parties’ vote shares change relative to each other.

For example, you might start with the SNP on 50%, Labour on 24%, the Conservatives on 15%, and the Liberal Democrats on 8%; then later have the SNP on 45%, the Conservatives on 30%, Labour on 15%, and the Liberal Democrats on 5%. With UNS you simply subtract 5% from the SNP vote, 9% from the Labour vote, 3% from the Liberal Democrat vote, and add 15% to the Conservative vote, in every seat. (And get negative votes for some parties sometimes.)

But the truth is that it’s more complex than this. That’s what ‘relative swing’ takes care of, by taking into account the ‘multiplier effect’. With the above numbers you multiply the SNP vote by 0.9, the Labour and Liberal Democrat votes by 0.625, and the Conservative vote by 2.0 in every seat. (And sometimes get a vote total over 100%.)

So my model is a mix of both, and currently tuned on 80% UNS and 20% ‘relative swing’.

The first step as described above will sometimes deliver negative votes in a few cases. And in some seats the vote total doesn’t add up to 100%. The second step is to eliminate all the negative votes. This is done simply by automatically switching from the model parameters to ‘100% relative swing’. This being a multiplier, it guarantees there can’t be any negative results. Third and last step is to deal with seats where the vote total is not 100%. This is also done automatically by recalculating votes proportionally from what the second step delivers.

Is the model foolproof? Of course not. No statistical model is, especially when you consider tactical voting. That’s why we can get wrong projections even with accurate polls (see ‘Hits and misses’ in my previous article). But I’ve compared my results with those you can get using Electoral Calculus or ScotlandVotes and they’re pretty similar. Not identical, but pretty close (more on that later). So I guess their underlying algorithms are pretty similar to mine.

How do I feed the model?

The source data are all the polls that I can find. Ideally I would rely on full Scottish polls (that is polls fielded in Scotland only and with a sample size of over 1,000). There were plenty such polls before the 2015 General Election and they proved accurate, unlike UK-wide polling. In the graph below the small dots are individual polls. The large dots at both ends are the actual 2010 and 2015 results. The trendlines show the evolution of vote shares and how polls correctly predicted the SNP landslide.


We are not that lucky this year. The snap election took everyone, including the pollsters, by surprise and there are many fewer full Scottish polls: nine so far and only five in 2017.


I therefore also have to rely on Scottish subsamples of UK-wide polls. A subsample size is typically 100 to 150, so they have a much larger margin of error (MOE) than full polls. This is the reason why many people think they should be discarded from any analysis. In principle I agree with this point of view, but this year subsamples are the bulk of the data we have. And there is a simple way to deal with the larger MOE and still get valid projections when interpreting the data. Which I will explain in the next section.

Continue reading General Election 2017: which way will Scotland go?

Predicting Scottish Councils elections: an impossible task?

By Jeremy Blackwell

A few words about me

Jeremy BlackwellJeremy Blackwell, almost 26 now, known on Twitter as @WeAreThe59. My friends usually call me Jez or JB but Jeremy is OK—actually I like it better. Born in France and studied statistics and political science there.

Living in Craiglockhart since late 2015 with my French boyfriend Alban and our beloved dog Bosco. Also member of the SNP, so no ambiguity about where I stand. Started seriously working on polls and projections during the 2015 General Election campaign; loved it and carried on during the 2016 Scottish Parliament campaign. And now giving a try at the 2017 Scottish Council election.

Hits and misses

Psephology is not an exact science. Sometimes polls are wrong. Sometimes your method is wrong. It’s closer to ‘trial and error’ than to rocket science, whatever the pundits might tell you about it.

Like everyone I totally missed General Election (GE) 2015 because UK-wide polls were so off the actual result. But I got Scotland right because Scottish polls were really accurate. Got the 56-1-1-1 part right; only got the exact seats wrong. I thought the Labour seat would be Glasgow North East; not Edinburgh South. I also thought that the Tory seat would be Berwickshire, Roxburgh and Selkirk, not Dumfriesshire, Clydesdale and Tweeddale. Still, I got the Lib Dem seat right. That was the easiest one.

I did better on Scottish Parliament (SP) 2016. Everybody got it wrong because the polls were wildly overestimating the SNP vote. But I was the ‘least wrong’ of all, crediting the SNP with only 66 seats while others went up to 70 or even more.

Now to the heart of the matter

The main question is: can Scottish Council elections be predicted with the same level of accuracy as Westminster or Holyrood elections? And the answer is a resounding no. Standard or sophisticated statistical methods that work for Parliamentary elections mostly fail for Council elections because there are many more variables that interact in sometimes unpredictable ways.

The first factor is that there are many more independent candidates in Council elections than in Parliamentary elections, and they get a much bigger share of the vote:

  • 0.1% in the 2015 General Election in Scotland
  • 0.3% in the 2016 Scottish Parliament election
  • but 11.8% in the 2012 Scottish Council election

The second factor is that Council elections definitely follow a wholly different pattern from other elections, like Westminster (WM) and Holyrood (HR) elections did before 2015 when they finally aligned. Council elections are still very far from aligning on the WM/HR voting patterns. One reason is that we have 59 MPs, 73 constituency MSPs but 1,219 Councillors after the last Boundary Review. So personal factors have a heavier influence on the Council vote, while Parliamentary votes are closer to a divide purely on party lines that can be statistically projected.

The third factor (and probably the most important) is the voting system: STV (Single Transferable Vote). There are many technical explanations of STV to be found (just Google it and make your pick). All of them are technically right but all are wrong on one major point. STV is not and has never been PR (Proportional Representation). Quite the opposite, in fact. STV is the perfect vehicle for tactical voting, and is even worse in many cases than FPTP (First Past The Post). Partly because FPTP is basic and simple while many people find STV extremely complicated and don’t fully grasp its limits and implications, and also how it can be used to maximise tactical voting.

Continue reading Predicting Scottish Councils elections: an impossible task?