Professor Charles Pattie writes widely on electoral issues, and is an author of two books on recent Scottish politics, including a study of the 1997 devolution referendum. He is also a Scot and is still British…..
Professor Charles Pattie’s previous post ‘The Scottish Independence Referendum: Home thoughts from not really abroad’ is available to view here.
After almost two years of campaigning, arguing and thinking, Scottish voters have spoken: by a clear majority (55% to 45%), they have voted against independence. The campaign, always lively and widely reported north of the border, suddenly came to life for people in the rest of the UK in the final weeks, in large part because since the start of September almost all the polls showed a dramatic tightening of the race. For the last fortnight of the contest, it seemed the result was too close to call. Yet, in the event, the result largely echoed what the polls had been telling us fairly consistently up until the last, feverish days of the campaign. Polls through to the end of August had pointed to a 10 percentage point No lead, and so it proved. So what happened? And what happens next?
What happened: how did the No campaign dodge the bullet?
Perhaps our first puzzle is to understand why the very tight contest indicated by the polls in the final weeks was not (quite) reflected in the actual result: why did we end up back where we started at the start of the summer with a 10 percentage point lead? If I am honest, the answer is not yet entirely obvious (though as more data comes in, the answer should become clearer). But two possible explanations seem likely, either alone or together.
First, much of the media narrative so far has focussed on the extent to which Alex Salmond’s dominance of the second televised leader’s debate, and even more so the sensational YouGov poll on the 5th of September which put the Yes campaign ahead for the first time, actually galvanised the No campaign. Prior to then, it had been widely criticised for being flat-footed, negative and passionless. But, with the risk of losing the referendum suddenly very real, the No campaign sprung to (relative) life. Gordon Brown emerged as a leading voice, pushing the No campaign to make a clearer statement that before on promising to increase the powers of the Scottish parliament in the event of a No win, and (even more dramatic) injecting a substantial degree of passion into the No campaign, not least in a barnstorming speech at the No campaign’s last eve-of-poll rally. A late surge by the No campaign might have been enough to widen the gap again by ensuring that committed No voters turned out and waverers switched to the No side.
But a second possible explanation also bears investigation. To understand it, it is worth remembering (if you can) the 1992 UK General Election. In the run-up to that contest, the incumbent Conservative government had faced a very bumpy ride, thanks to the poll tax, emerging disputes within the party over the EU, Mrs Thatcher’s defenestration as leader, the economy tipping into recession, and so on. For months before the contest, opinion polls had suggested the upcoming election would be close, and that a hung parliament was likely. In the event, that did not happen. The Conservative vote share in the actual election was almost unchanged compared to the previous contest in 1987 while Labour’s share edged up but by only a small amount. The reason? It seems many Conservative supporters knew how they would vote but (given strong negative public reactions to policies like the poll tax) were deeply reluctant to say so publicly: instead they told pollsters they were undecided. As a result, almost all the polls substantially under-estimated the real level of Conservative support. It is at least plausible that something similar was happening in the final weeks of the referendum campaign. Momentum seemed to have shifted to a noisy and increasingly self-confident Yes campaign which presented itself as speaking for Team Scotland, by implication – sometimes subtly, sometimes more overtly – suggesting that the No camp was anti-Scottish (never a popular position north of the border). In such a climate, it would not be too surprising if many No supporters decided to keep their heads down until polling day. As in 1992, this would lead to the polls under-predicting their numbers.
What happened: Dissecting the results
What of the referendum itself? What lessons do the detailed results hold? Several features really stand out. Let’s focus on two in particular: turnout and the geography of the vote.
The referendum really did put paid to the persistent myth that the public is politically apathetic. Levels of involvement in the campaign have been very high indeed. Meetings have been well attended, people have actively sought out information, have engaged in debates with family, friends, strangers – this really has been an extraordinary event (just as a marker of that, I was in Edinburgh on the last weekend of the campaign, attending the annual conference of British psephologists: walking around the streets of the city I constantly accidentally overheard snippets of conversation about the decision facing Scottish voters – a thing I have never experienced before in almost 30 years of studying elections).
That extraordinary degree of public engagement kept up right to the end of the campaign. Something like 97% of the Scottish electorate registered in the run-up to the referendum, an extraordinarily high rate. At 84.6%, turnout was remarkably high. To put that in context, no other election in modern British political history has seen such a high level of voter engagement: we have to look back to 1950 and 1951 to find other election turnouts over 80%. The highest turnout on the night was a truly extraordinary 91% in East Dunbartonshire, followed closely by a turnout of 90.5% in East Renfrewshire (both are dominated by some of the more comfortable of Glasgow’s middle class suburbs). The lowest turnout, meanwhile, was Glasgow’s 75%. But to put this into context, at the 201 UK General Election only 10 constituencies out of 650 recorded a higher turnout than Glasgow’s referendum vote. We have never seen such levels of public engagement in a British vote.
Why? Two simple reasons. First, we know that turnout tends to be higher in contests where voters think there is a lot at stake than in contests where they can see few real differences between the rival protagonists. And here was a vote which really mattered. What is more fundamental than the future of one’s country? Second, we know that voters turn out in larger numbers when their vote really matters. In conventional elections, most of us live in safe constituencies, where the result is a foregone conclusion: our vote doesn’t matter. Only a few live in marginal seats, where a few votes changing from one party to the other could change the result. This is why marginals are where the parties concentrate their campaign efforts. In consequence, one criticism of the current system for Westminster elections is that it means the campaigns do not even try to reach most voters living in safe seats, hence the generally low turnouts we have seen at recent elections.
But this contest was different. It was a simple national poll (so no safe or marginal seats). And over the final weeks poll after poll indicated a knife edge contest. Under those circumstances, every vote in every part of the country counted, and a high turnout was only to be expected.
That said, turnout did vary from place to place. The graph below compared the percent who voted in the independence referendum in each local authority (the smallest areas for which referendum results are available) with the percent turnout there at the 2012 Scottish local government elections. As we might expect, there is an obvious positive relationship: turnout at the referendum was highest in communities where it is usually relatively high, lowest in communities where it is usually relatively low. The striking thing isn’t the relative pattern of highs and lows (that is predictable). The surprise is the very high level of turnout.
That said, it is important to keep this in perspective. We are not (at least not yet) sitting on the brink of a new era of mass involvement in politics. Don’t bank on a big rise in turnout at the 2015 UK General Election. The core turnout lesson of the independence vote is that the referendum was sui generis, very much a unique and unrepeatable event. Once again, this was a contest which really mattered. On the result hung an existential choice: whether a 300-year old state continued or came to an end. The contest was close. And every vote counted, everywhere. These are perfect storm conditions for high turnout. Come the next UK general election, however, the choices, important though they may be, will be more mundane and will have the get out clause that if we don’t like the results we can always vote for the other lot next time. And at general elections, most of us do live in safe seats where there is little or no chance of affecting the result or of being contacted by any political party. We may not be as politically apathetic as some think us to be. But nor are we so politically engaged that we will participate even of the stakes are low and/or we have little real chance of hanging the outcome. Do not expect the 2014 Referendum to usher in a new era of high mass participation. Rather, expect turnout at the 2015 General Election to be not dramatically different from turnout in 2010.
And will the high levels of engagement engendered by the referendum campaign last in other ways apart from electoral participation – for instance, in the form of an upsurge in grassroots movements? Maybe. But again it would pay to be a little sceptical until clear evidence comes in. While some of those energised and engaged by the referendum, especially on the Yes side, may become politically active in other campaigns too, others will feel let down and marginalised. After the euphoria of the campaign, and the disappointment of the result, many will abandon political participation once again. A new era of public engagement in politics? Perhaps – but don’t count your chickens yet.
The geography of the vote
Even more than in a General Election, in a referendum it is the headline figure which really matters: 55% of Scotland’s voters opted to stay in the UK, while 45% opted for independence. But that masks considerable local variation. The diagram below gives the results by local authority, ordered from the local area with the lowest Yes share (Orkney, with 33%) to the authority with the highest (Dundee, where Yes took 57%). Scottish local authorities vary hugely in size, from Behemoths like Glasgow (with almost 490,000 voters) and Edinburgh (378,000) to minnows like Clackmannanshre (40,000), Orkney (18,000), Shetland (18,500) and the Comhairle nan Eilean Siar (the Western Isles Council to you, with 23,000). The diagram therefore appears to exaggerate the overall No lead, as small authorities are given the same visual weight as large ones. But several patterns still stand out. First, after all the sound and fury of the campaign, only four local authority areas returned a Yes majority – and strikingly they were all urban central belt councils. Dundee has been an SNP stronghold for very many years, so a Yes win there is hardly a surprise. But what of Glasgow, North Lanarkshire and West Dunbartonshire? And in Inverclyde, the race really was neck and neck: No prevailed by a margin of only 0.16 percentage points, wafer thin by anyone’s standards. What is striking is that these are the industrial, working class West of Scotland heartlands of Labour’s long post-war hegemony in Scotland. The SNP has been making vigorous inroads there in recent local government and Holyrood elections, but even so the extent to which voters there have embraced independence must be a source of considerable trepidation for Labour especially.
Elsewhere in Scotland, however, No came out on top, even in areas like Aberdeenshire, Angus, Moray and West Lothian where the SNP has been electorally strong down the years (in a telling moment on referendum night, Alex Salmond, leader of the Yes campaign, decided not to attend the count in his Aberdeenshire home but to go instead direct to the national count outside Edinburgh). As was widely anticipated, the Borders, Dumfries and Galloway and the Northern Isles all proved strong opponents of independence (these, after all, had been the local authorities whose populations had been least enthusiastic about devolution in both the 1979 ad 1997 referendums). But so too were middle class areas like East Renfrewshire, East Lothian, East Dunbartonshire, Edinburgh and Stirling.
In general, as the next graph shows, the more support the SNP had locally at the 2012 local government elections, the better the Yes campaign did in 2014. But the relationship is not overwhelming. The geography of Yes support was not a simple case of the SNP mobilising its support base. Also important was the Yes campaign’s capacity to tap support from disillusioned Labour supporters.
More telling, however, is the comparison (shown in the next graph) between the geography if the Yes vote in the independence referendum and the Yes vote 17 years earlier, in the 1997 devolution referendum. The relationship is very strong (indeed, it is by quite a margin the strongest of any I look at here): the greater the support for devolution in a local area in 1997, the greater the independence vote in 2014. The places where an appetite for more control of Scottish affairs by Scottish politicians is strong have remained the same over time: what has shifted is the extent of control demanded.
The results reveal splits between urban and rural Scotland, between working class and middle class communities, and between the more remote areas and the Central Belt (especially around Glasgow). These variations are not, of course, new. Nor are they unexpected. They are evident in most Scottish elections, and were certainly there at the 1997 devolution referendum But at that earlier contest, every local authority returned a Yes majority: the only question was how large. That level of unanimity across the country is clearly lacking in the independence vote.
Another factor worth quick consideration is the link between turnout and support for independence. There was much speculation in the run-up to the referendum that a high turnout, by empowering many first time voters (both young voters and older people who had never previously taken part in an election) would benefit the Yes campaign. At least at an aggregate level, however, this does not seem to have been the case on the night. As the next graph shows, the relationship between turnout in a local authority and the Yes vote share there is weak and negative. In other words to the extent that there is any relationship between vote and turnout, a higher turnout seems to have favoured the No camp. Evidence of a late surge to save the Union? Probably not: remember that turnout was highest in places where it always tends to be highest, and these are the more middle class communities of Scotland, where support for independence tends to be weakest: the relationship between turnout and support for independence is probably an accidental by-product of the relationship between each of these and class.
What happens next?
So what now? Scotland’s voters have spoken, a majority opting to stay in the UK. But what now? Does this mean we will return to the status quo ante, as though nothing had happened? I suspect not. One of the ironies of the referendum is that it has locked the UK into a period of radical change, whatever the result. Clearly, the most radical change on offer, the break-up of the Union, is now off the table. But a clear promise (a ‘vow’, no less) has been made by the leaders of the three main No parties regarding establishing greater tax and spend powers for the Scottish parliament in the event of a No vote. That result has now been delivered: the Westminster party leaders must now deliver in their turn, or they risk feeding a sense of betrayal among Scottish voters. And that can only end one way – in a future and much more successful Yes vote.
The Conservatives, Liberal Democrats and Labour have publicly committed themselves to a new settlement for Holyrood. But the details remain not only obscure but also contested. The three parties leaders made a public vow in the final days of the referendum campaign (published on the front page of Scotland’s largest tabloid newspaper, the Daily Record) that this new settlement will include greater tax and spend powers for a Scottish parliament, and retention of the contentious Barnett formula which has guided the distribution of government funds to the constituent parts of the UK since the late 1970s, and which, controversially, delivers more government funds per head to Scotland than to England. So the Scottish Parliament is in effect being told that it will be allowed to raise much more money, to spend it on more things, but that its already generous rate of subsidy from the UK taxpayer will be safeguarded – hardly a fast track to fiscal rectitude.
But the exact provisions of this new dispensation for Scottish government are less clear. Despite their leaders’ vow, the three parties which dominated the No campaign differ substantially in their proposals for new powers. For instance, Labour’s proposals (involving an extension of the Scottish government’s tax-raising powers) are much less far-reaching than the Lib Dems (who wish to move to a fully federal system) or (ironically) the Conservatives (who propose full fiscal devolution).
And the Labour and Conservative leaderships also have to sell their proposals to their own parties – and that will not be easy. It was Labour MPs from the North East of England who hamstrung that party’s devolution plans in the 1970s by adding a clause to the Scotland Act 1978 requiring that devolution would require not just the support of a majority of voters in Scotland, but also the support of at least 40% of registered electors (the Yes vote reached the first threshold, but not the second, meaning devolution failed: the fallout brought the Callaghan government down, ushering in the Thatcher years). What made those MPs in effect sabotage their own party’s legislation was a deep unhappiness with a proposal that would give powers and resources to people north of the border that were denied their own constituents. Similar motivations may well run now: Labour MPs representing depressed communities in England may well resent the extra resources and powers now being promised to Scotland. And do not forget that many (all?) Labour MPs still believe strongly in a universal welfare state, in which benefits are the same for all: more power to Scotland undermines that, especially if future Scottish governments gain control over the pensions and benefits system there (currently this is a policy area for which responsibility is reserved to Westminster), as it makes it more likely that benefit rates and pension payments would vary from one part of the UK to another. They might – not unreasonably – see such a move as a body blow to the idea of universal welfare. Will they let this go without a fight?
Similarly, many Conservative MPs are very unhappy indeed at their leader’s acceptance of the ‘vow’. The agreement to retain the Barnett formula is already proving unpopular among Conservative (and some Labour) MPs, for instance. Within hours of the referendum result being called, for instance, James Wharton, Conservative MP for Stockton South, wrote an article for the influential Conservativehome website (see http://www.conservativehome.com/platform/2014/09/james-wharton-mp-i-cannot-ignore-the-threat-that-devolution-poses-to-my-region.html) voicing his concern that this would disadvantage his constituents: he is not alone. For others, more powers moving north of the border further undermines the Union.
The leaders’ vow to give more powers to Scotland also raises the ante for the UK’s other devolved governments. Wales, for instance, is much poorer than Scotland (the 3rd wealthiest region of the UK), but the Welsh Assembly has fewer powers than the Scottish Parliament and thanks to Barnett, Wales receives less public spending per capita than Scotland does. Not surprisingly, Welsh politicians are already arguing that increasing powers to Scotland while guaranteeing Barnett formula further entrenches Welsh disadvantage. Calls are being made for a similar guarantee of more powers and resources for Cardiff. Things are more complicated, inevitably, in Northern Ireland. But here, too, a guarantee of more powers for the Scottish Parliament raises worries. Expect to see, in the near future, more powers for the Welsh and Northern Irish assemblies.
The biggest challenge of all is what to do about England. With around 84% of the UK population, it is by far the largest part of the UK. But so far it has not been well served by movements towards devolution elsewhere in the UK since 1997. Most fundamental of all, more powers being passed to the Scottish Parliament also pushes the West Lothian Question (WLQ: in its crudest form, why should Scottish MPs vote on issues which do not affect their constituents) even further up the political agenda. Almost overnight, the WLQ has gone from a concern of constitutional experts and political anoraks to a mainstream issue. Even during the final week of the campaign, several leading Conservatives (including John Redwood MP and Boris Johnson) spoke out publicly against the position that their party leader was adopting, and demanding English control over English matters. Whatever new settlement merges, it will involve some degree of devolution to England or to lower-level bodies.
But quite how that will be achieved is harder to see. Some advocate a devolved English parliament with significant control over policies affecting England alone. However, this will be difficult to do in practice, not least because England is so very much larger than any of the other constituent nations of the UK. An English parliament would, inevitably, dominate the UK’s political landscape. Its decisions would inevitably have major consequences for the rest of the country (because of spill-over effects, implications for the distribution of resources from one part of the UK to another, and so on), and to a much larger degree than their decisions would have an impact on England. I can think of no federal or quasi-federal political system anywhere in the world which has anything like the same level of asymmetry in population between its largest state and the rest. That is not, I am sure, accidental.
If not an English Parliament, then what about special reserved business in Westminster, where only English MPs would be allowed to vote on English issues? This has been widely canvassed by the Conservatives, and is their preferred option. However, it is also fraught with more difficulties than you might imagine. First, it creates two classes of MPs: English MPs and the rest. Second, how easy will it be in practice to define purely English issues? Remember the huge population disparity between England and the other parts of the UK. To take just one complication, decisions on many English policies would almost inevitably have spending implications which would affect the distribution of resources between the constituent parts of the UK. At that point, Scottish, Welsh and Northern Irish voters – and hence their MPs – might not unreasonably claim to have some interest in the discussion, as it might affect the resources available for their own devolved bodies to work with. Or to take another example, if English MPs were to vote for substantially different trading standards to those adopted elsewhere in the UK, that could have a huge impact on business practices and on trade between the constituent parts of the UK – and once again would be something MPs from the other parts of the UK might have a reasonable interest in discussing. Once again, Scottish, Welsh and Northern Irish MPs would have a case for arguing they should have a say. What is more, EVEL (English Votes for English Laws) does not have cross-party support. While the Conservatives favour it, Labour does not, fearing they might at some future date find themselves with a Westminster majority thanks to the presence of Scottish and Welsh Labour MPs, but with no majority among English MPs. They would then find themselves in the constitutionally remarkably difficult position of being the UK government but being unable to get the bulk of their legislation through because it would be voted down in English MP-only sessions (the same challenge is much less likely to face future Conservative governments, for obvious reasons). What might on the face of it seem a simple and neat solution, therefore, could in practice quickly become rather messy.
For the political anoraks, the McKay Commission, set up in early 2012 to look at how the House of Commons might deal with the WLQ, offered a partial solution in its final report (see http://webarchive.nationalarchives.gov.uk/20130403030652/http:/tmc.independent.gov.uk/). Under their proposals, all bills coming before the Westminster Parliament would have to declare which parts of the UK they affected. All Westminster MPs would vote on legislation, but ‘decisions at the United Kingdom level with a separate and distinct effect for England … should normally be taken only with the consent of a majority of MPs for constituencies in England…’. The key word here is ‘normally’, which allows a get-out clause should some English measure be felt to have large spillover effects for the rest of the UK. How this might work in practice and how easy it will be to explain to the public, however, is anyone’s guess.
Or perhaps English regional government? Maybe: Yorkshire, for instance, has a distinct regional identity (and then some), and a population comparable to Scotland’s. And other large countries (Germany, for instance) have very successful federal arrangements which devolve significant power down to similarly-sized areas across the entire county. A fully federal option is favoured by the Liberal Democrats and by some in Labour. The catch, however, is whether there is substantial public demand in England for such a tier of government. At the moment, there is not.
What is more, a full-blooded federal system would entail substantial devolution of tax and spend powers on a wide range of issues. There is already pressure (discussed above) to devolve responsibility for welfare benefits to the Scottish Parliament, and some models envisage that body taking control of raising and spending most of Scotland’s tax revenue. That would be a radical step in its own right, undermining the universal welfare state. Extend it more widely to new regional bodies throughout the UK and the universal welfare state is undermined even further. Pensions and benefits payments, for instance, could vary substantially from one part of the country to another, as could tax regimes. And the closer we get to each region taking control of most of its tax revenue, the harder it becomes to run cross-subsidies from the richer to the poorer parts of the UK. Areas with higher needs would increasingly have to meet those needs from their own relatively limited resources. Goodbye the post-war Welfare State. Some may oppose this, and some may welcome it. My point, however, is that even though No won, the fallout of the 2014 Independence vote could push us very rapidly in this direction without much prior thought or debate.
Another thought. The closer we move towards a truly federal system in the UK, the greater the pressure will become for a modern written constitution, carefully delimiting the powers and responsibilities of the different levels of government. Goodbye to the UK’s fabled and flexible ‘unwritten constitution’. But since the current confused state of affairs facing us following Scotland’ independence referendum is a direct consequence of our piecemeal and ad hoc approach to constitutional reform, tinkering with different bits of the UK’s governance structure without carefully thinking through the implications for other parts, maybe a more formal codification of the constitution would be no bad thing.
On a more positive note, one unequivocal triumph of the referendum was the extension of the franchise to 16 and 17 year olds, who engaged in the debate and vote with evident relish, energy and maturity. Now they have had a vote on the most momentous political choice they are ever likely to face, it becomes very hard to justify not extending the same rights to 16 and 17 year olds for other elections, including those for Westminster and Holyrood. Whether this can be done in time for the next UK General Election in 2015 is not clear. But I suspect it is no longer a question of if full voting rights will be extended to 16 year olds, but only of when.
So if you were thinking that the No vote marks the end of the matter, think again! The status quo is no longer an option. More powers for a Scottish Parliament must come, and they will have to be quite substantial: to renege on that now would fuel an even larger and this time potentially unstoppable independence movement. And that in its turn makes it more, not less, likely that further devolution will occur elsewhere in the country, especially in England. And that has implication for things like universal welfare, the scope for resource transfers from rich to poor regions, a written constitution for the UK, and so on. The constitutional genie is out of the bottle: it isn’t going back any time soon.