the politics of Tidy Britain why Labour lost and how we win again By Chris Bryant MP

Transcription

1 the politics of Tidy Britain why Labour lost and how we win again By Chris Bryant MP

2 the politics of Tidy Britain why Labour lost and how we win again Chris Bryant s perspective on the 2010 election is a welcome contribution to the Smith Institute s project on progessive politics and policies for a fairer society. His thought provoking analysis shows how the political landscape is changing, what that means for policy making, and why it is important to recognise the reasons for the collapse in Labour s core vote. Chris highlights the key lessons from recent elections and offers a new approach to win back Labour voters. We would like to thank Chris for this excellent piece of work. Paul Hackett Director, Smith Institute Published by the Smith Institute This report represents the views of the author and not those of the Smith Institute. The Smith Institute September 2010

3 Contents Chapter 1: Repairing Labour s fractured core Chapter 2: The scores on the doors 2.1 National heartland trends 2.2 Heartland seats 2.3 Key in-play marginals 2.4 Ethnic diversity in safe seats 2.5 Importance of heartland voters to future success Chapter 3: Campaign for Tidy Britain Conclusion: Labour s relations with Tidy Britain Author s biography

4 Chapter 1 Repairing Labour s fractured core 3

5 Repairing Labour s fractured core It is in the nature of things that politics always tends towards complacency, and at a first sight the Labour Party could gently pat itself on the back at the moment. Sure, we are not in power and with the government secure in its majority in both the Commons and the Lords, barring mishaps there is no reason why there should be another General Election before May 2015 (or 2014 if a shorter 4 year fixed term parliament proposal is agreed). That could be a very long 54 months. But, say the eternal optimists, the swingeing and excessive cuts that the government are set to introduce will alienate not just key public sector workers and those who most depend on public services, but in risking a second recession may well anger a far wider part of society. Moreover the Liberal Democrats, tied to the Tories like human body armour, will be flayed alive in successive local, Scottish and Welsh elections. Seats will tumble into our lap. Considering what might have happened, so the theory goes, we are in a remarkably strong position to come back. Some, who share this perspective, go further. They argue that Labour s best chance of success at that distant General Election lies in re-enthusing our core vote so that disillusioned former Labour voters return to the fold. Reasons not to be too cheerful The trouble is that such an approach risks consigning Labour to a worse defeat in Of course we have to be honest about our failings during 13 years in government. Of course there are elements of what we did that we need to expiate if we are to re-attract some of our historic coalition. We have to be clear-sighted in our analysis of where in government we adopted policies out of pollster-led blindness or media-driven anxiety regardless of the facts, or out of slavish defensiveness following the unexpected loss in But we would be completely delusional if we were to pretend that we can win a general election if we were to abandon some of the essentially correct insights that inspired Labour s recovery under Kinnock and Blair. Most importantly, we have to start by being accurately and acutely honest about how bad a result we had this year. The relatively healthy number of seats we secured (258, precisely the same as in 1959, when we won 44% of the vote) masks some worrying facts and trends. 4

6 We secured just 29.7% of the vote, doing worse than at any modern election other than the disastrous 1983 election when we got a mere 28.2%. The 6.5 percentage point drop in the Labour share of the vote since 2005 is the worst fall in any one general election cycle, again with the sole exception of 1983 (down 9.5 percentage points). The cumulative drop in the share of the vote over the last two elections (down 12.3 percentage points) is worse than that of 1979 and 1983 (down 12.0 percentage points). This was the first time that Labour s share of the vote has fallen for three elections in succession. We lost more seats (93) in one election than ever before since Labour is now in third place or worse in 213 seats including 139 of the 197 seats in the East, South East and South West regions. The comparable figures before the election were 135 and 95. In particular we lost heartland seats that have returned Labour MPs for decades like Redcar, Burnley, Dewsbury, Crewe and Nantwich and Nuneaton. In the South of England outside London we won only 10 out of the available 212 seats. In Wales Labour held 26 out of the 40 seats, but achieved just 36.2% of the vote, our worst result since % lower than we got in We now have just 4,809 councillors across the country, compared to the Tories 9,446, representing a fall of 5,799 since The numbers increased very slightly this year, but the Tories still hold 46% of all council seats. Likewise, in terms of political control of councils, Labour increased its take this year to 54. But the Tories hold 202 councils, gaining 181 since By contrast Labour now has control of 152 fewer councils than in 1997 and at different points has lost control of Bassetlaw (lost 2004), Crawley (lost 2006), Dudley (lost 2002), Hammersmith and Fulham (lost 2006), Harlow (lost 2002), Burnley (lost to NOC in 2003 and won by LD in 2010), Redditch (lost to 5

7 NOC in 2004, won by Con in 2008), Liverpool, Newcastle (lost to LD in 2002), Birmingham (lost to NOC in 2003) and Bristol (lost to NOC in 2003 and won by Liberal Democrats in 2009). These results, and the trends behind them, should alone point to a larger problem than a complacent strategy could possibly address. In addition, though, the future is strewn with new electoral hurdles that should steer us away from complacency. Thus the government s plans for constitutional reform may hit Labour (and the Liberal Democrats) hard. Even if the government does not explicitly gerrymander the boundaries of seats at the next election (as draft legislation would allow them to do) new, larger constituencies based on electoral registers with (or rather without) 3.5 million missing voters could cut the number of inner city and working class seats. This could disproportionately affect Labour, as could a change in the law regarding the electoral quota for Welsh seats. If the alternative vote referendum is carried (which I suspect it will) and implemented (which may be another matter) the next General Election could be on a completely different, and quite possibly unknowable basis. So it is all the more important for Labour to combat complacency of every kind. We should not afford ourselves any comfort zone neither the intellectual comfort of simply defending our recent record nor the emotional comfort of returning to the politics of pure oppositionalism of the 1980s. We will not win the next election by default. The whole premise of the government s deficit plan is based on being able to make popular tax cuts in the immediate run-up to a 2015 election. We cannot presume on the collapse of the Liberal Democrats, nor should we assume that if the Liberal Democrat support subsides it will deliver votes or seats to us. Indeed it may well give succour and a majority to the Tories. After all, of the Liberal Democrats 57 seats, the Conservatives are in second place in 38 and Labour only in 17. For that reason Labour now has to be remarkably clear-sighted in its analysis of the 2010 General Election. We need to prepare to fight the next election, not the last, or indeed the 1997, 1992 or 1983 elections all over again. We need to eschew the nostrums of both New Labour and Old Labour, not because our arguments were wrong then, but because they will almost certainly be outdated in In the words of R H Tawney, whose analysis following the 1931 defeat rescued the Party from almost certain permanent ignominy, we need to engage in the intolerable burden of thought, because the only effective brand of politics is one that constantly reassesses itself. 6

8 Reading the runes Many others have already analysed why we lost in 2010 (not least the candidates for the Labour leadership), but the tendency has been just to review what happened in the marginal constituencies and amongst marginal floating voters. This is a mistake. It has skewed the analysis. Of course it is not enough to win our safe Labour heartland seats like the Rhondda. We may not need to take the constituencies of Windsor or Chelsea and Westminster to win an election, but we do need to reach into Kent and Hertfordshire and regain a swathe of support across the South of England and the Midlands if we are to win the next General Election. Indeed we need quite specifically to win the popular vote and the largest number of seats in England. Likewise it will not suffice simply to rely on winning over disillusioned Liberal Democrat voters and adding them to our safe pot. In addition to re-energising former Labour voters who either didn t vote at all this time round (or indeed in 1997, 2001 and 2005) we will need to win over people who voted Conservative this time. We therefore must be a movement of true electoral ambition, reaching out to the vast majority, if we are to aspire to government again. But we also have to be honest with ourselves on five key counts. First, there is no such thing as safe any more. In common with other parties we have found more and more voters choose rather than inherit their political allegiance. Even in the Rhondda (which has returned a Labour MP for a hundred years and was the only seat in Britain in 2005 to return an MP with more than 40% of the electorate, including non-voters) more people say I vote Labour or I m voting Labour this time than I am Labour these days. In part this is a natural function of increased educational opportunity and a general pick and mix approach to most aspects of modern life. But the implications for Labour have already been extreme and, as noted above, have seen us lose seats at parliamentary and council level that should be part of Labour s most reliable heartland. In addition I would suggest it is actually inappropriate to consider many voters as floating or swing as they now effectively consider themselves independent, not because they intend to vote for candidates who have no party allegiance, but because they have no ultimate party allegiance themselves and intend, if they vote, to support the most attractive person or party at the time. Asking what independent voters think should be a constant part of the task ahead. Second, over the 13 years of Labour government the bedrock of Labour s support fractured. Much has been made of the relative collapse in Labour s support amongst 7

9 socio-economic groups C1, C2, D and E. Some have rightly suggested that we need to re-energise the D and E vote, much of it disillusioned by Labour s supposed failure to deliver for it. But while it is certainly true that many working-class voters were disillusioned and refused to vote Labour, my strong conviction is that the 13 years of Labour government fractured the historic Labour vote in a far more radical way, allowing a wedge to develop between the aspirations of Britain s low-paid workers and the needs of those who have relied on benefits; between the civil rights defenders of the middle classes and those whose main concern was their own security; and between the indigenous and the migrant working class. It was not just that one group was alienated from the other. A profound sense of injustice began to develop, with tidy Britain, whether working working class or working middle class, often resenting what they saw as their scrounging neighbours and those on benefits resenting a country in which the wealthy seemed to prosper disproportionately. At its most acute this was expressed as a deep anger about immigration and, separately, the needs of the black and minority ethnic population. At its mildest it left pensioners complaining when they were not entitled to a benefit that those who had worked all their lives were penalised whilst those who had never worked a day in their life got all the benefits they wanted or needed. For the Tories, it was an opportunity to campaign against Iain Duncan Smith s broken Britain that tidy Britain so heartily disliked. Third, the debate on whether Labour needs to reach out to its core working class support or to middle Britain is based on a false dichotomy. In fact during the leadership campaign the media has been far keener to fashion this dichotomy than any of the candidates, unfairly characterising one candidate as a core vote candidate, another as a middle England-er. Of course the bedrock Labour DE vote exists to a greater or lesser extent in every marginal seat in the land, and the fact that our DE support dwindled to varying degrees across the country at the last election needs to be addressed. That is necessary. We cannot win marginal seats like Crawley or Stockton without securing the Labour bedrock in those constituencies. Necessary but not sufficient. As important as it is, when the Government is adopting a cuts from cradle to grave approach, to renew our commitment to delivering for the DE vote, without an accompanying determination to deal with the legitimate political perceptions of the tidy working class and the middle classes, our efforts will fail. Fourth, we have at times, through our policy decisions and language, made the mistake of thinking of our core vote as a homogenous block with united views. We have presumed that the needs and values of C2s are closely aligned with those of Ds and Es. The 8

10 middle classes have also been mistyped as wholly divergent from the aspirations of C2s a point not helped by the likes of the multi-millionaire semi-aristocrat David Cameron ludicrously declaring himself middle class. But the truth is that the median income in the UK is not 100,000 or 60,000 but just over 20,000. The failure of Labour to garner sufficient support amongst C1 voters, let alone C2s, was a central determining factor in the election outcome and it will be again, come the next election. 44.3% of the electorate are white working class voters, in work and not claiming benefits and a further 25% are lower middle class living in their own homes. They represent the two groups amongst whom our vote fell furthest at this election a massive swathe of tidy Britain we need to win back. Fifth, an increasing number of working-class C2 voters, many of whom turned against us this time round but voted for us in 1997, 2001 and 2005, have seen their outlook on life radically transformed. They take foreign holidays, they eat out, they buy designer clothes things their parents didn t even think of aspiring to. Often aspiration far outstrips opportunity (this is as true for politicians as it is for working-class voters) but Labour has to face the fact that an increasing number of those whom it historically relied on for support undoubtedly see themselves as part of an aspirational class, whatever their income or their job. They want to get on in life. They want to do better, be better off than their parents and they dream of their children doing better even than them. They work hard and hope for a comfortable retirement. They don t look to the state to provide everything for them, indeed they often aspire to enough financial independence to be able to opt out of state provided services and go private, not just for their gym, their golf club or their transport to work, but even their doctor, their dentist and their children s school. When we are seen or perceived to undervalue or curtail those hopes and ambitions they have understandably abandoned us. In some cases voters have expressly felt that, having succeeded in life, Labour quite simply isn t for someone like them. Reasons to be careful So, without reconfiguring the heartland and bedrock vote Labour stands little chance of returning to government, but without seeking to reach out to the hopeful, the ambitious, the successful and the rich Labour will never be able to deliver the prosperity based on justice and fairness that we avow. Four caveats: first, no analysis of an election is perfect. It is difficult to allow for highly personal issues such as individual public responses to the leader or the leadership team, especially against the background of the growing number of independents. Sociologically virtually identical constituencies can and do behave 9

11 very differently, either because of local issues or in response to particular candidates. Moreover, despite the best efforts of pollsters and psephologists, the precise reason anyone votes in a particular manner remains unknowable. Thus, for instance, one factor in the election is extremely surprising, considering the lengthy anti-mp campaign waged by several national newspapers and the expenses debacle: incumbency seemed to play a key positive part in several Labour seats. The average swing against Labour was actually 2% lower where the sitting MP was standing again. Seven seats (Dudley North, Halifax, Kingston upon Hull North, Newport East, Plymouth, Moor View, Tooting and Wakefield) were won by Labour with narrow margins in which incumbency may have been decisive, whilst we narrowly lost 11 (Brighton, Kemptown, Brighton, Pavilion, Bury North, Carlisle, Halesowen & Rowley Regis, Nuneaton, Pudsey, Sherwood, Thurrock, Warrington South, Weaver Vale) where the Labour MP had stood down or retired (often late). Second, we should not allow any one election victory or defeat overly to determine our strategy for the future. The shock of losing in 1992 made us learn our lessons too well. Aware that we had seemed too anti-success and dogmatic in our advocacy of state-based solutions in opposition in the late 1980s and early 1990s, in government we leant over backwards to be seen as pro-business and not anti-city. It meant we were too reluctant to regulate the City against its own market failures, too hesitant about tackling grossly over-optimistic lending. It meant that when it came to the credit crunch we were too hesitant to criticise the banks or crack down on the City s excesses. It also made us too slow to deal properly with the issue of posted workers, often working under oppressive conditions and undermining British workers terms and conditions. That does not mean that we should abandon the key precept that social justice and prosperity, a strong and a fair economy are not mutually exclusive, but mutually dependent. Nor does it mean we should ditch a commitment to wealth creation and economic growth as the powerhouse of regeneration, but it should point us towards a more active strategy for growth, including export driven growth that requires a manufacturing regeneration. Third, we cannot afford to renege on all that we stood for through 13 years of government. The Tories and Liberal Democrats have very determinedly set about trashing our achievements. They say we did nothing to reform our outdated constitution. Yet we brought devolution to Wales, Scotland and Northern Ireland. They say we undermined civil liberties. Yet we completely transformed the landscape of LGBT equality and brought in the Human Rights and Freedom of Information Acts. They say we bankrupted the country. Yet our action in the midst of the global 10

12 economic crisis prevented a far worse recession or slump. So we should not, in our honesty, trash our own record. Fourth, although voters dislike parties that appear out-of-touch or dogmatic, they also seem increasingly reluctant to support parties whose platform is entirely confectioned to fit the pollsters analysis of the voters prejudices. Principles and policies that appear consistent with those principles are not the icing, but the cake itself. Without self-evident principles no party can expect to weather the storm of traditional or internet media analysis. All too often a popular policy is sustained by little other than warm air and when it cools the politician is left holding a very deflated balloon. We constantly need to explain how our historic values match modern concerns, but there is no better platform Labour can adopt than that which it holds to be true. 11

13 12

14 Chapter 2 The scores on the doors 13

15 The scores on the doors It is undeniable from an analysis of the voting trends in Labour s 50 safe seats and the national trends of core Labour demographics, that there has been a detachment of heartland working class voters from the Labour Party. This detachment began and in many cases, was most pronounced at the 2005 General Election, but has been reinforced by this year s results. Labour s safe seats have suffered significant drops in the majorities that were enjoyed in 1997 and While in most cases the dramatic loss of vote share seen at the past two elections is insufficient to really weaken Labour s hold on these safe seats, the trend calls into question our ability to reach out into marginals that are crucial to our future success where the heartland vote is strong. Similarly the general pattern in our key 2010 marginals was that Labour did no better in fact marginally worse in seats where heartland C2 and D demographics made up a significant proportion of the vote. We lost a number of seats where stronger support from these demographics would have been decisive in us keeping them. Looking to the future, in the 39 seats that Labour could win by overturning a majority of 5% or less at the next election, C2, D and E groups feature heavily and will be as important demographics to reach out to as ABC1s. Regaining support amongst heartland demographics will also be key to holding onto our super-marginals. While the general trends point towards a disastrous election for Labour amongst aspirational and low skilled workers and the unemployed and welfare dependent, there are positive stories amongst other equally important heartland demographics that Labour must not alienate if we are to stop the rot. Most notable amongst these are black and minority ethnic voters (essential to our relative success in London safe seats and marginals) and older people (nationally the trends for older people are not particularly positive for Labour but in safe seats in Scotland, they constitute a significant proportion of our heartland vote). 2.1 National heartland trends National trends show that C2DE groups are an increasingly important battleground between Labour and the Tories at a national level. While support for the Tories amongst the middle classes has stagnated since 1997 (2010 levels were exactly the same as 1997), support has grown by 10% amongst C2DEs, and attracting these voters was central to relative Tory success this year. 1 These gains correlate to a decline in Labour support amongst C2DE groups over the past two elections in particular. 1 All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, researchspecialisms/socialresearch/specareas/politics/trends.aspx. Data are estimates based on campaign polling and exit polling. All references to vote share or swing amongst C1,C2,D,E voters are taken from Ipsos Mori data. 14

16 For Labour, the loss of working class and welfare dependent voters has had a far greater impact on our vote share at a national level than loss of middle class voters since Labour has lost 7% support from middle class ABC1 voters since 1997, but this is far overshadowed by a combined 40% loss of vote share from C2DEs. It is these voters that the Tories fought most keenly for in reaching out to Labour marginals this year, and it is the loss of these that has contributed most to Labour s decline in votes if not seats since Over the course of the past four elections, Labour has consistently performed better amongst DEs than C2s. This was true in 1997, when we had the support of 59% of DEs to 50% of C2s, and remains true in 2010, where we got 40% support amongst DEs to 29% support amongst C2s. We have also seen a slightly steeper decline in support amongst C2s than DEs - having lost 21% of our vote share amongst the former since 1997 in comparison to 19% amongst the latter and in 2010 we lost 3% more support amongst C2s than amongst DEs. Figure 1: Labour s C2 and DE vote 70% 60% 50% 40% 30% 20% C2s DEs 10% 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, researchspecialisms/socialresearch/specareas/politics/trends.aspx. The fracture in the working class Labour vote So the trends in support from C2s and DEs are both on a downward trajectory and taken together, these voters have shifted their support away from Labour to a far greater degree than middle class voters. Does this then not point towards adopting a core vote strategy? 15

17 I would argue no. We must not make the mistake of assuming that mutual decline in support amongst C2s and DEs means that these two sets of voters are moving closer together and voting more alike. In fact, the opposite is true. The core vote as we knew it in 1997 has fundamentally fractured. The aspirational working class (C2s) have shifted their voting trends further away from those of the DE poor since 1997 and have done so to the extent that their patterns almost exactly mirrored the voting trends of lower middle class C1 voters in The first graph below (figure 2) compares the trends in C1 and C2 votes from 1997 to In C1/C2 levels of support for Labour and Conservative over these four elections the trend is clear these sets of voters have gotten closer together. The second graph below (figure 3) compares the trends in C2/DE votes from 1997 to In contrast, levels of Labour support amongst these two groups show a general trend of moving further apart, with a bigger gap registered between the two groups from 1997 to Figure 2: C1 and C2 voting trends 60% 50% 40% 30% 20% 10% C1 LAB Support C2 LAB Support C1 Tory Support C2 Tory Support 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, researchspecialisms/socialresearch/specareas/politics/trends.aspx The difference between Labour support amongst C1 and C2s has shrunk from 13% in 1997 to just 1% in Similarly, the difference between Tory support amongst C1s and C2s has shrunk from 10% in 1997 to just 2% in The difference between Labour 16

18 support amongst C2 and DEs however has widened slightly since 1997, and is now at 11% in C1s and C2s have converged to such an extent that at the 2010 election they represented a virtually unified voting block, with DE voters demonstrating fundamentally different, stand alone behaviour from that of the aspirational working class and the lower middle class. As both graphs demonstrate, the convergence of C1 and C2 votes and the pattern of increased fracture between C2 and DE votes, are long term trends that gathered pace after the 2001 election represented an acceleration of these trends on what we d seen between 2001 and 2005, when long term frustration amongst the aspirational working class collided with a viable alternative in the Tories. Figure 3: C2 and DE voting trends 70% 60% 50% 40% 30% 20% 10% 0% C2 LAB Support DE LAB Support C2 Tory Support DE Tory Support Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, socialresearch/specareas/politics/trends.aspx C2s have, therefore, fractured from the core vote and now share the values and voting trends of C1s not DEs. This presents a challenge for the Party in positioning ourselves with traditionally Labour voters. Going forward the Party must not make the mistake of clumping aspirational working class voters together with the low skilled, unemployed and welfare dependent poor under the assumption that a policy approach that gains the support of one group will automatically gain that of the other. C2 voters are demonstrably frustrated, disillusioned and disengaged with this approach and, in 2010, have fundamentally shifted support in reaction to it. 17

19 Where is the fracturing most pronounced? The fracturing of the working class vote was most pronounced between C2 and DE women in 2010, when the gap between these two groups more than doubled to 20%. While women DEs have consistently been Labour s strongest heartland demographic (including in 2010 where 45% of women DEs voted for us), we saw a dramatic fall in Labour s support amongst C2 women. Support amongst C2 women was down 15% this year on 2005 levels to 25%, marking a 22% decline in support from this group on 1997 levels. 2 In contrast, there was a much smaller degree of fracturing between C2 males and DE males in 2010 than the 20% that now separates C2 and DE women. In fact, 2010 brought the two male groups much closer together, and they are now separated by only 2% (C2 male support was 33%; DE male support was 35%). The fracturing has narrowed from 9% in 1997 and 8% in 2005 because of a sharper decline in DE male support than amongst C2s the opposite trend to the women. We have lost 25% of the support that we had from DE men since 1997, which was at 60%, while the decline in male C2 support was 18%. It is therefore the shift in voting trends amongst aspirational working class women that contributed most to the fracturing of Labour s heartland vote, and the loss of support from both these women and low skilled and welfare dependent men that had the biggest impact on Labour s overall vote nationally in Figure 4 shows the relative decline in support amongst C2s/DEs and the different subgroups of C2s and DEs since Where have our lost heartland voters gone? Our loss of support amongst C2 women and DE men has been the Tories gain. The Tories have picked up a 13% bounce amongst DE men since 1997 and 12% bounce amongst C2 women. The bulk of both of these increases 8% and 7% respectively happened in 2010, pointing towards a crystallisation of the shift to the right amongst key heartland voting groups this year. 3 While Labour has suffered from a longer-term decline in support amongst DE men and C2 women, it wasn t until 2010 that these groups of voters positively endorsed the Tories on a large scale. The long-terms patterns amongst male DEs and female C2s are shown in the figures 5 and 6. 2 Ibid 3 Ibid 18

20 Figure 4: Labour s relative decline of C2 and DE votes 70% 60% 50% 40% 30% % 10% 0% C2s DEs Female C2s Female DEs Male C2s Male DEs Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, researchspecialisms/socialresearch/specareas/politics/trends.aspx. Figure 5: Male DE voting trends 70% 60% 50% 40% 30% 20% Male DEs (LAB) Male DEs (CON) 10% 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, researchspecialisms/socialresearch/specareas/politics/trends.aspx. 19

21 Figure 6: Female C2 voting trends 60% 50% 40% 30% Female C2s (LAB) 20% Female C2s (CON) 10% 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, com/researchspecialisms/socialresearch/specareas/politics/trends.aspx. More broadly, support for the Tories amongst C2s and DEs as whole groups has been on an upward trend since 1997 as support for Labour has declined. In 2010, support amongst C2s for the Tories hit 37%, and significantly, for the first time since 1987 the Tories achieved greater vote share amongst C2s than Labour. Much of this increase in C2 support came from a surge in support amongst women C2s, 41% of whom voted Tory to 25% Labour in 2010 (in 2005 it was 40% LAB 34% CON). Support amongst DEs for the Tories while on an upward trend, increased to 31% in 2010, still 9% lower than Labour support amongst this group. The difference between C2 and DE Tory vote in % to 31% - was, at 6%, the same as it was in However, this represents a slight narrowing of the gap from In 2005, following surges of 4% support amongst C2s for the Tories in both that and the previous election, C2 support stood 8% higher than DE support. In 2010 however, it was DE support for the Tories that surged with a 6% increase, and as a result, the gap reduced back to 6%. 20

22 Figure 7: C2 voting trends 60% 50% 40% 30% 20% 10% Labour C2s Tory C2s 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, com/researchspecialisms/socialresearch/specareas/politics/trends.aspx. Figure 8: DE voting trends 70% 60% 50% 40% 30% 20% 10% Labour DEs Tory DEs 0% Source: All raw data taken from Ipsos Mori, How Britain Voted 1997, 2001, 2005, com/researchspecialisms/socialresearch/specareas/politics/trends.aspx. 21

23 Both C2s and DEs registered surges in support for the Liberal Democrats in 2005, increasing their vote share by 4 and 5% respectively. This trend was continued amongst C2s in 2010, when they increased their support for the Liberal Democrats by 3%, to 22%. However, support amongst DEs stagnated (in fact fell very slightly by 1%), so that the gap between C2s and DEs for the Liberal Democrats was at 5% in 2010, wider than in the past four elections. Support for the BNP also increased amongst both groups at a national level, with DE support increased by a quarter to 12%, and C2 support increasing by 4% also to 12%. The 2010 Watershed The fracturing of the C2 and DE vote is growing bigger, and represents a fundamental shift in the definition of the traditional core Labour vote. While Labour support amongst both C2s and DEs has markedly declined since 1997, the pace of decline has been more pronounced amongst C2s where voting patterns more closely reflect those of the lower middle class than DEs. This has led to a widening of the gap in Labour support amongst C2 and DEs to 11% in 2010, up from 8% in This fracturing has been particularly noticeable amongst women. Taken together, C2/DEs have gradually offered growing support to the Tories since However, 2010 marked a significant shift both for DEs and particularly for women C2s. For both groups, 2010 constituted the first positive endorsement of the Tories on a large scale. In the case of C2s, 2010 was a watershed moment for the Tories as they beat Labour. We had lost these voters gradually but this year they were offered a real alternative in the Tories, and this crystallised the shift away from the Labour Party. 2.2 Heartland seats The slump in Labour support amongst C2 and DEs has translated into a significant downward trend in levels of support for Labour in our safe seats. As with the national trends amongst C2 and DEs, the decline in support in our safe seats cannot be isolated to 2010, but is a longer-term reflection of the detachment that traditional working class Labour voters feel towards the Party, and this year, has developed into a more positive endorsement of the Tories. It is misleading to look at the 2010 safe seat results in isolation, and doing so gives a skewed picture of where Labour stands in our heartland seats. For instance, across our 50 safe seats, the average vote share for Labour actually went up in 2010, very 22

24 marginally, by 0.024% on 2005 levels. 4 However, this reflects more the disastrous results in these seats in 2005, which the Party managed narrowly to bounce back from this year, rather than a positive endorsement of Labour. A longer-term analysis is required to get a proper impression of the trends in our safe seats. If we set 2010 results in these seats against the results in 1997, Labour has dropped, on average, 12% of our vote. (NB. It is only possible to do a direct comparison in 32 seats because of boundary changes and the creation of new seats). Labour has seen a drop in vote share in all but two of these 32 seats. Our average vote share across the 32 in 1997 was 69.6%; this fell to 57.4% in The trend is particularly pronounced in Wales where all safe seats have dropped between 17% and 27% on 1997 levels. 5 Credit where it s due Labour did make gains in pockets. There were positive swings to Labour in 4 London seats against the Tories, and against the Liberal Democrats we made gains in 3 London seats and 5 seats in Liverpool. Most strikingly, there were positive swings in 12 of the 16 Scottish seats where our opponents were the SNP. In these seats, Labour achieved a positive swing of 1.95%. All in all, in 2010 Labour achieved a positive swing in 26 of our 50 safe seats, with an average vote share increase on 2005 of 5.95% across these 26. We owe much of what we did achieve in many of these seats to an excellent ground game, played out by an army of dedicated volunteers. However, again, setting the results seats where we gained votes in 2010 in the context of 2005 (and not 1997) belies the longer-term downward trends of Labour support in our safe seats. While some voters did return to Labour between 2005 and 2010, in 26 of our safe seats the overall pattern is still one of decline from 1997 levels. In the 14 seats where it is possible to compare Labour s positive swing in 2010 to the 1997 result, these results represented an average decline in vote share of 7.9% over the 13 years. This is despite these 14 averaging a 5.4% vote share increase between 2005 and Therefore, while some results in our safe seats point to a return to Labour for some of our heartland voters in the short-term, the trend from 1997 is still a significant loss of support. This longer-term slump in our heartlands gathered pace in In 23 of the 32 seats where direct comparison of vote share between 1997, 2005 and 2010 was possible, the result in 2005 was worse in terms of lost vote share than in In only 8 seats was 4 Raw data taken from BBC Election 2010 website 5 All 1997 election result data taken from Parliamentary Research Paper, General Election 1997, Table 18 hclibrary.parliament.uk/rp2001/rp pdf 6 All 2005 election result data taken from BBC Election 2005 webpage map/html/map05.stm 23

25 2010 actually a worse result. This reinforces two conclusions: 2010 was not an isolated result but fits into a broader downward trend; and the loss of support is therefore largely not the result of factors unique to 2010, such as the personality of leaders or the impact of events, but longer term disengagement with the party amongst heartland voters. Where has the heartland vote gone? Broadly speaking, Labour s decline in our safe seats came alongside a decline in support for the Liberal Democrats. Voters in our heartland seats defected from the centre left to the centre and extreme right. Our main opposition in safe seats varied pretty evenly between the Tories, Liberal Democrats and SNP. We fought the Tories in 14 seats dispersed across London, North East and West and Wales. We fought the Liberal Democrats in 17 contests in London, North East and West and Wales. And we fought the SNP in 16 Scottish seats. Labour lost a greater proportion of voters in safe seats where we were up against the Tories, than seats where we were facing a challenge from Liberal Democrats, SNP or Plaid Cymru. In the 14 safe seats where we fought the Tories, the average swing was 3.15% to them. This is in comparison to the 17 seats where Liberal Democrats were placed second in which the swing was an average of 0.9% to them, and the 16 seats where the SNP was the primary challenger, where Labour achieved an average positive swing of 1.95%. We did make gains against the Tories in four London seats. However we lost ground to them in the other 10, scattered across the North East (three), North West (three) and Wales (two), with one in Yorkshire and one in London. The swings in the North West were generally slightly smaller than those in the North East and Wales. Generally speaking, the Tory increase in support in this group of seats was a new phenomenon in In 7 of the LAB/CON seats, the Tories overturned a loss in vote share in 2005 to a gain in This trend was most pronounced in seats like Jarrow, in the North East, where the Tories overturned a loss in vote share of 0.5% in 2005, to increase by 7.8% in The pattern reflects the increased appeal of the Tories to heartland voters between 2005 and The Liberal Democrats The picture in our safe seats where the Liberal Democrats were our main opponents was far more balanced than in the LAB/CON seats. The average swing was much 24

26 lower as noted, and while the Labour vote share decreased by 2.3%, the Liberal Democrats only picked up an average 0.24% of this. In nine of these 17 seats, Labour made gains. Our gains came in London and Liverpool where there was a notable decrease in support for the Liberal Democrats. In the North East and in Wales there were sizeable swings to the Liberal Democrats, though in some cases this was in spite of a drop in the Liberal Democrats vote share. Indeed, if you look at the 11 seats which marked a swing from Labour to Lib Dems, the Liberal Democrats only outperformed the Tories in terms of improved vote share in two seats Easington, where the Liberal Democrats gained 3.1% in vote share to the Tories 3%, and Sheffield Brightside and Hillsborough, where the Liberal Democrats 6.9% to the Tories 1.7%. In Swansea East for instance, the overall swing was from Labour to Lib Dems of 1.7%, but the swing was brought about by increased Tory vote share (of 4.7%) rather than a positive endorsement of the Liberal Democrats (who saw their vote share decrease by 1.8%). This pattern was reflected across a number of ostensibly Labour to Lib Dem seats. The overall conclusion to draw from this is that in these heartland seats, as in the LAB/CON seats, the trend was of traditionally centre-left voters defecting to the Tories. SNP In Scotland, the Liberal Democrats suffered the biggest losses. The overall pattern (12 of the 16 seats) was of a swing from the SNP to Labour. However, in seven of these 12, the SNP actually made gains, and in almost all 12 seats, the SNP either lost a smaller proportion of the their vote share than the Liberal Democrats, or gained, even in cases where the overall result was a swing in Labour s favour. BNP Whilst election night seemed to deliver a rebuff to the BNP, thanks especially to the results in Barking and in Dagenham, the BNP has actually made significant inroads into 25 of our 50 safe seats generally saw them increase their vote share or register significant vote shares in seats where candidates stood for the first time. A more detailed analysis of the BNP s performance is discussed later in the chapter. Labour is undergoing a long term slump in our support amongst heartland voters in safe seats. In 2010, while the picture was in some seats an improvement on terrible 2005 results, the results broadly served to entrench this trend of decline in support and saw the Tories and the BNP take advantage of our disconnected heartland voters. 25

27 2.3 Key in-play marginals This analysis is based on the trends in the seats that, pre-election, Labour held with majorities of between 5-10% (but not, therefore, those with majorities of less than 5%). The overall picture in these key in-play marginals was of a 5.5% swing in support from Labour to Conservative, with a slight dip in Lib Dem support also contributing to the Tory increased vote share. 7 As with the pattern in many safe seats, this year s result compounds the downward trend in key marginals beginning in 2005, where there was an average 3% swing from Labour to Conservative. 8 Viewed against where the Party was in these seats in 1997, the long-term decline is far steeper than that evident from 2005 to Between 1997 and 2010, in the 52 seats where comparison is possible, Labour s vote share has dropped by 14.4% (8.3% of that drop in vote share happened between 1997 and 2005). 9 Regional effects In terms of how the decline in support for Labour in 2010 translated into marginal seats held and lost, the picture varied significantly from region to region. Of the 18 in-play marginal seats we lost, the areas where our decline in vote translated into most seat losses were: the East Midlands, where we lost three of our four in-play marginals and saw an average decrease in our vote share of 8.15%; the Eastern region, where we also lost three out of four and lost 6% of our vote share; non-london seats in the South we lost four of seven and suffered a 8.2% swing; and Yorkshire, where a 7% decrease translated into losing three of six seats. The positives in our in-play marginals came in London, where despite a 3.7% swing to the Tories, we retained our 5 marginal seats; in the North West, where a 6.9% drop in vote share was not sufficient to have a real impact on our seats we lost only two of 13; and in Wales despite a 4.5% swing to the Tories and a significant increase for the BNP, we held all five. 7 Raw data taken from BBC Election 2010 website 8 All 2005 election result data taken from BBC Election 2005 webpage flash_map/html/map05.stm 9 All 1997 election result data taken from Parliamentary Research Paper, General Election 1997, Table 18 hclibrary.parliament.uk/rp2001/rp pdf 26

28 Demographic trends in marginals The demographics in the in-play marginals we lost offer some clues into what some of the key factors were in us losing them. The 18 seats that we lost have higher proportions of white population than the marginals we held: amongst the 18 the average percentage white population is 96% compared to 93% in the seats we held. 10 The 18 seats also have on average higher levels of aspirational working class and low skilled workers than in our holds (in the 18, C2s and Ds constitute 15% of the population in comparison to 13% in the holds). 11 In contrast, levels of unemployment and long term welfare dependency do not appear to have impacted on Labour support in the marginals. Unemployment was at 4% in both our holds and losses, 12 and there was only a 0.3% difference in levels of long-term JSA and Incapacity Benefit claimants in our holds vs losses. 13 So while we might conclude that higher levels of C2 and D voters in the marginals looks to have had a negative impact on Labour s ability to hold onto a seat, support amongst E voters shows no such impact on Labour s success in the marginals. There were a number of marginals that we lost where loss of support from aspirational and low skilled working class voters looks to have been particularly decisive. In Erewash for instance, where 20.4% of the population are in C2 and D jobs, we lost the seat with a 10.5% swing to the Tories. In Amber Valley, C2 and Ds make up 18.4% of the population and here we lost the seat following a 9% drop in vote share. In Keighley, C2 and Ds make up 18.6% and we lost it with an 8.3% swing. Similar trends occurred in Carlisle and Gloucester. An analysis of Labour s performance across all of the 55 in-play marginals not just those that we lost - based on portions of C2 and Ds also points to the impact that this voting block had on Labour s success. Across these 55 marginals, the median percentage of the total population that are C2 and Ds is 13.9%. In the 26 marginals that have above that median level (i.e. have higher proportions of C2/Ds) the average swing away from Labour was 6%. In contrast, the swing in the 27 seats below the median was a percentage point lower. In the seats with higher proportions of C2 and 10 Raw ethnicity data for each constituency sourced from Office of National Statistics (ONS), Census Raw employment data for each constituency sourced from ONS nomisweb using data from Annual Population Survey 12Raw unemployment data for each constituency sourced from ONS nomisweb using data from Annual Population Survey 13 Raw data on JSA claimants sourced from ONS nomisweb Labour Market statistics 27

29 Ds, the swing away from Labour in 2010 was a marked increase on the 2005 swing, which had been on average 2.5% Labour Conservative. Labour performed worst and ultimately lost marginal seats where C2/D voters were most prevalent. There were of course local issues and broader considerations that influenced these races, but a trend is visible and supported by the national analysis of what happened to Labour s support amongst C2s in particular. Older voters The patterns in support amongst older voters differ considerably between safe and marginal seats. In our safe seats, the older vote was absolutely crucial to our success. Labour has 17 safe seats in Scotland, all of which have 33% pensioner populations or above. Outside of London and Liverpool, we achieved our best results in Scotland, and an average vote share gain of 5.5%. The Tories (and in some cases SNP) also made gains in these constituencies, at the expense of the Liberal Democrats who saw almost universal drops in vote share. In contrast, in our marginals, Labour fared better in younger seats. In the 27 seats with below average pensioner population, Labour suffered a 5% swing to the Tories, but in the 28 older constituencies, the average swing away was 8%. Older voters were also an important constituency in our marginal North West seats, where 10 seats had 18.5% or more pensioner population. While there was a 4% swing away from Labour across these seats, we only lost two of them. 2.4 Ethnic diversity in safe seats A chunk of Labour s safe seats have significant black and minority ethic populations and in these seats, Labour achieved our most positive results. Labour s safe seats can be broken down into three broad categories 14 in relation to ethnicity: Category 1: ethnically heterogeneous seats where the white population varies from 34.5% to 67.9%; 14 These categories have been set simply on the basis of logical groupings of seats on the basis of ethnicity i.e. there are three distinct blocks that, in my view, it makes sense to group together. Of course it is contestable that seats at the top end of category 1 begin to in fact constitute category 2 seats, but in my view, the gap between category 1 and 2 is significant enough (67.9% white pop. to 83.1%) to conclude that these seats have fundamentally different levels of diversity. 28