Sunday, 21 September 2014

"What the Scottish independence referendum showed us"; my view

In the aftermath of the Scottish referendum I sat down to write the definitive post-mortem of the result. After a few minutes I stopped. Why? Because we have data from 32 councils. That's it; 32 councils. So it was a little surprising to see Steven Fisher running with a piece describing "What the Scottish independence referendum showed us". In the piece he uses data from the 32 council areas to draw conclusions about the particular groups which had voted Yes or No. He makes two broad points: first, that "economic hardship led people in the hardest hit areas to conclude that economic management from Westminster was not working for them." And second that "support for independence was considerably lower in those places where more of the voters were born elsewhere in the UK."

I respect the work Mr Fisher does a great deal. In many ways he goes about things in a different way than I do, but that's what draws me in. I am not stupid enough to think my way is the right way. I keep an open mind. However I don't think it's safe to use data from just 32 councils to make the claims he does. It's a simple statistical condition that analysing small samples is unsafe. Let's not forget, in the weeks running up to the referendum we had academics lining up [rightly] to pick apart the samples from pollsters. It would be hypocritical to now use a sample size of 32 to reach apparently definitive conclusions about the independence referendum.

Which brings me to my last point; data. In my view the independence referendum has given us election nerds very little to chew on. The opinion polling was sporadic and at times unreliable; there was no exit poll; the results were announced by council area. All of which means that it's pointless to even begin to analyse what happened, at least from my perspective. I have around eighty variables for each of the 32 council areas (at the last count!): age, gender, ethnicity, place of birth, economic activity, household composition, etc., etc., However what is the point of even beginning to run models? The spatial unit of analysis (council area) is far too large to deliver meaningful results. I suspect there'd be a relationship between some form of deprivation and the Yes vote but at what scale? And who can say how the "deprived" voted? You can say that a council area like Glasgow has pockets of deprivation. You can also say that Glasgow voted for Yes. But you can't say that "the deprived voted Yes", certainly not from the data we have.

So, after all that, I am sorry to say that the data from the Scottish referendum does not allow ME to say what happened in any great detail. Others can try, and have. Unfortunately I don't have the data I want.

Thursday, 18 September 2014

Labour, UKIP and the referendum

I have been conducting research into the demographic groups which have been attracted to UKIP from Labour in England and elsewhere. One particular group which does this is older voters in social housing or former right-to-buy properties. These voters vote Labour but a healthy chunk of them have been attracted to UKIP in recent local and European elections, but also further back in other elections.

With this in mind I have produced a map showing where these voters are in Scotland, overlaid with the boundaries of the 32 council areas for tonight's IndyRef count. Here it is:

Let me be clear here. I am NOT saying that ALL these voters are going to vote UKIP, that would be preposterous. Nor am I saying that ALL these voters are going to vote for Labour. That's not what this map is about.

This map shows the spatial distribution of those voters who, elsewhere in the United Kingdom, have shown they are dissatisfied with Labour and have used UKIP as a means of expressing that dissatisfaction. Whether they are going to use a Yes vote in the independence referendum as an alternative means of kicking Labour and the Westminster establishment is a moot point, but I thought I might at least put it up.

Wednesday, 17 September 2014

Scottish referendum maps

So, here are some maps. There are two 'types' of maps. Where possible I try to use standard deviation maps as they allow easier comparison. Essentially the warmer the colour the stronger the particular variable.

Friday, 12 September 2014

Heywood & Middleton

First of all, let me say that I will be in the seat a few times between now and polling day to get a flavour of things, and I'm happy to meet with anyone from the media to talk about the seat demographics, etc. Drop me a line using the blog contact form and I'd be happy to help. But, on with the profile.....

Following the untimely death of popular local MP Jim Dobbin Heywood & Middleton will hold a by-election to replace him on October 9th, which is the same date as the by-election in Clacton. I suspect this is not a coincidence; however more of that later. First, as always, the demographics:

Before I go on I want to compare this profile to that for Wythenshawe and Sale East, another by-election which took place in February:

Now I am going to put up the demographic profile for Clacton:

And finally I want to add the coefficients between these demographic groups and UKIP vote share in the 2014 local elections:

You will see, first of all, that the profiles of Wythenshawe and Heywood & Middleton are very similar, whereas the profile of Clacton is somewhat different. In broad terms what these profiles show is that Clacton is largely made up of older, probably comfortably retired groups whereas both Heywood & Middleton and Wythenshawe possess far higher proportions of working-age, blue-collar or struggling poorer voters. You will remember that in Wythenshawe UKIP devoted a good deal of effort to draw support away from Labour in one of its perceived bastions and secured second-place with 18% of the vote, about where I expected them to be.

However there are much larger proportions of blue-collar working voters in Heywood & Middleton than there were in Wythenshawe. Around four in ten of voters in Heywood & Middleton are blue-collar households (I43 to K50) compared to just two in ten in Wythenshawe. Take a look at the correlations above. In 2014 UKIP has done very well with these precise groups. UKIP has done similarly well with the poorest voters too (O67 and O69s). They are strongly represented in Heywood and Middleton, as they were in Wythenshawe.

This is an introductory profile but I want to make one point before I go. Clacton is all but won for UKIP. It certainly can't take anything for granted but the strength and direction of travel of polls is fairly conclusive. But the Heywood & Middleton by-election comes two weeks after their party conference in Doncaster, a conference they are apparently using to challenge Labour in its back yard. If they're true to their word, and if they possess the resources, they might reflect that Clacton can be left to its own devices and that Heywood & Middleton is a better use of its limited resources; after all, Labour wouldn't want that. And in politics it sometimes pays to do whatever your opponent least wants you to do.

That said, I expect Labour to win the seat and for loyalty to Mr Dobbin to play a large part. However for junkies like me I want to see whether UKIP can make very serious inroads into Labour in a bona-fide Labour heartland.

Wednesday, 10 September 2014

Scottish independence; Survation poll from 10th September

Survation today published its poll on the Scottish independence referendum. They have included crosstabs for the regions of Scotland. The map below shoes the regional poll results overlaid on to a constituency map showing the winning party in the 2010 Westminster parliamentary election. Further down is the same map, this time overlaid with the results from the 2011 Scottish parliament elections. And finally is the same map, but with the results from the 32 local authorities for May's European election results. This last map is important because the results next week will be announced by each of the 32 local authorities as and when their count is complete.

Tuesday, 9 September 2014

The Empty Nesters

Whenever I grow weary of parenting (yes, there are such days!) I only need to reflect that one day the nest will be empty. It's a sobering thought but for many of us there will come a time when the house will not be filled with the happy laughter, stamping feet and crashing cutlery of children. We will be empty nesters. Freed from the responsibilities of family life we will no doubt adapt to changed circumstances in the same way our parents have.

All those thoughts, and more besides, crossed my mind when I began to research this blog on the electoral behaviour of empty nesters, more particularly those empty nesters which are moving away from the Conservatives and, increasingly, voting UKIP. Between 2005 and 2014 some of these empty nester households, particularly the more affluent ones, have drifted slowly away from the Conservatives and found a home with UKIP. Not that the Conservatives are losing them all; many of them are still loyal to the party. The maps below shows the geographical distribution of Conservative-UKIP leaning empty nesters. On the left is the standard deviation maps showing the proportion of households in each constituency (green = low, red = high). On the right is the cluster map showing those areas with statistically significant clusters of empty nester households. These are mostly affluent or at least comfortable households.

You will see that Clacton is among the seats with high concentrations of these groups. There are also higher concentrations in seats like Amber Valley in Derbyshire, which should concern the incumbent Conservative MP Nigel Mills, sitting as he does on a 1.2% majority over Labour; seats like Great Yarmouth and Waveney in the east where UKIP might expect to perform strongly; seats like Mid Dorset and North Poole in the south; seats like Thanet North and Thanet South on the Kent coast. All these seats have strong proportions of empty nester voters, many of whom have voted for UKIP as far back as 2005.

An important point to remember regarding empty nesters is their turnout record. It's hard to find any demographic grouping which has a stronger record of high turnout in all elections, whether they are local, European or national. These voters are strongly civic-minded and believe it is important to do your duty come election time. They are likely to have voted Conservative for much of their adult life but in more recent elections they have been attracted to UKIP.

Much attention has been paid to the problem Labour have with UKIP. This is of course true. However there are still a fairly wide range of demographic groups which are moving from the Conservatives to UKIP. Empty nesters provide just one of those groups. Their geographical distribution, combined with strong turnout record, means that in certain seats they may hold the balance of power in 2015; particularly those seats where Labour and the Conservatives go head-to-head. A point of reassurance for local Conservative campaigns has been the differential turnout records of Labour and Conservative leaning voters. The assumption, until now, has been that older, more affluent [Conservative] voters turn out in higher numbers than younger, less affluent [Labour] voters. However in seats like the ones above a move to UKIP by older, more comfortable empty nesters might very well dampen this effect, and neutralise any perceived turnout edge for the Conservatives.

Saturday, 6 September 2014

The east coast: I can see Parris from here!

See what I did with the headline? I know; I should find someone else to write them in future. As somebody who has found enough favour to be granted the privilege of writing for a national newspaper I know only too well how a dramatic headline can misinterpret your original carefully calibrated comment piece. So when I saw the headline "Tories should turn their back on Clacton" above Matthew Parris's smiling face I cut him some slack. He didn't actually mean that, did he? Well, surprisingly he did. Or at least it said so in the piece I read!

Which got me to work. Suppose the Tories did decide to turn their backs. What would this look like? The map below shows the winning party of those seats on the eastern coastline in 2010. Next to it is a map showing the vote share achieved by the Conservatives in those seats, followed by the margin of victory for each of the winning parties.

There are 38 seats along this stretch of coastline from Cleethorpes on the Humber estuary to Portsmouth on the south coast. In 2010 the Conservatives won 32 of them, with an average vote share of 46%.

This is a remarkable table in some respects, and should concern Labour. Labour have lost 14 seats along this coastline since 1997. For the Conservatives these seats became something of a heartland, going as they did from 19 seats in 1997 to 32 in 2010, and increasing their vote share accordingly. Elsewhere I have noted that if one were to walk the entire coastline from the Humber estuary to Bristol (incl Cornwall and the south coast) you would walk through just a handful of Labour seats. This may or may not surprise some when one considers some of these seats are retirement destinations, and the party has performed poorly amongst older voters.

Which brings me back to Matthew Parris, who has rightly identified Clacton as a seat with higher than average older populations. This is undeniable, as I have shown in another post on Clacton. But which particular demographic groups are most prominent in Clacton, and what do they look like? Below is the demographic profile of the seat:

Two groups jump out, the L53s and L54s. These make up almost a quarter of all households in Clacton. These are retired people living by the sea in retirement communities or bungalows. These households have household incomes of around £20k per year and have some savings accrued from a lifetime in work and possibly equity and report they are coping well financially. They take two or more holidays a year.They are overwhelmingly Church of England and faith is really important to them. They believe strongly it is important to do your duty, and also do not like to be surrounded by people from different cultures or backgrounds. They experience very little crime personally but are concerned about it. They are also very likely to have left school at 15 or 16. These groups have typically voted Conservative since 1979 but in more recent times have been attracted by UKIP. They also have an extremely good turnout record.

These demographic groups are present up and down the eastern coast, but are most prominent in seats like Clacton, Norfolk North, Norfolk North West, Boston & Skegness, Waveney, Cleethorpes and Louth & Horncastle. This alone should concern CCHQ, but more worrying is the transmission of Mr. Parris's world view to seaside communities along the coastline. The worry for them should be that many of these voters have always voted Conservative, delivering seats to the party through thick and thin. However they might well choose, as they have in recent local elections, to take the opportunity to turn their own backs......this time on the Conservative Party itself.