Constituencies 2010-'19, Explaining Political change in Britain
Published: 09/01/2020 - by James Prentice. Cite: J.Prentice, "constituencies 201-'19: explaining political change in Britain, University of Sussex, working thesis 2020.
The purpose of this article is to highlight how the United Kingdom, and in particular England & Wales, has experienced substantial electoral change that has altered the parties’ support bases, and from this dramatically altered electoral competition between Labour and the Tories. Academically, this process is known as a "electoral realignment". This purpose of this post is also to outline the causes behind electoral change of constituencies within England & Wales across the last 10 years. The third purpose of this article is to identify how electoral competition has changed across the last decade in a more detailed way than Labour's “Red Wall” losses suggest is the case. This post's broader focus is vital as the Red Wall narrative has missed broader trends that tell a more detailed story of how UK politics changed during the 2010s.
This post argues the last decade of political change in Britain reflects an electoral realignment. It discovers clear trends where smaller towns outside the influence of universities, which have also been hurt by de-industrialisation, became more Conservative leaning, whilst more unban cosmopolitan areas became more likely to back Labour. It also finds that Brexit and demographics within a constituency became more important in determining voting patterns, whilst traditional economic class groupings have developed a distinctly different effect on Labour’s and the Conservatives Party’s vote share. For example, those in working class occupations have helped increase the Conservative's vote share, historically something which has not been the case. These are trends that extend far beyond Labour’s Red wall and instead tell a wider story about cultural change in Britain and reaction to such change has translated into new voting patterns.
In order to achieve these objectives this post will compare and analyse constituency data from the 2010 and 2019 election results. It will compare the result for both of the two main parties constituency by constituency and analyse trend changes in the share of the vote these two parties achieved. It will also outline trends in geographical changes that will show in more detail the extent and nature of electoral change. The article will then utilise models to determine what the causes electoral change might have been and then make conclusions on how this has impacted British politics.
Results of my research into constituency change 2010 - '19:
Long-term demographic trends and geographical changes:
The age divide has increased over the decade with passing elections. Younger voters have increasingly been diverting their support to Labour since the 2010 election. The data also shows that most of the Conservative’s largest decreases in support existed within constituencies that have higher than average amounts of younger voters. This only occurs in constituencies with very youthful populations, highlighting the stark divide between younger and older generations that has emerged across the decade. Importantly, this indicates that electoral change being influenced by generational factors is not limited to just the change between the 2017-19 elections, but it is a trend spanning across the last decade of UK elections. Oppositely, constituencies with older than average demographic profiles conversely displayed some of the biggest gains over the decade for the Conservative Party. The greater the amount of older voters the larger the Conservative Party gains tended to be, with Labour generally seeing larger decreases of support. This meant that in these constituencies the net gain over Labour was quite considerable. Significantly, this again displays the growing social divide and strengthens the potential that growing social divisions has led to a changed electoral arena for Labour and the Conservatives.
The qualification divide has grown greater with successive elections throughout the last decade. With regards to constituencies containing fewer people obtaining any qualifications the Conservatives have increased their level of support considerably, whilst Labour experienced losses. Therefore, the net gain the Conservatives achieved over Labour in these constituencies is around 15%, much greater than the average swing. This indicates that again the constituencies that most mirror sociodemographic divides, in this case those with and without qualifications, have resulted in the largest swings between the two main parties. Constituencies with a greater number of voters with higher level qualifications, such as A-level and above qualifications, have become more likely to return a Labour MP. The trend has increased throughout the decade, in 2017 Labour became more reliant on constituencies with a disproportionate number of voters with higher level qualifications. This trend strengthened in the 2019 election, indicating that social divides have grown throughout the decade. With Labour being the main beneficiary of support from groups with higher level qualifications this has seen their support rise in areas they traditionally have struggled to perform. Labour has lost out in areas with fewer people with degrees, meaning that they have lost support in some traditional Labour areas to the Conservatives throughout the decade. This might help explain why Labour has been able to increase their support in areas like Brighton and Canterbury, but at the same time not in traditional working-class Labour areas. Therefore, those with fewer qualifications have realigned themselves around the Conservative Party and those on the opposing side appear to have realigned around Labour. This again indicates how societal changes may have developed new voting allegiances.
The Brexit effect?
The change in the Conservative Party vote can partly be seen to have arrived through Leave seats they won. The Conservative Party’s share of the vote in the 2010 election within seats that would go onto vote to leave the EU was quite a bit lower than Leave seats they now hold in 2019. This indicates that across the decade they have been increasing their vote share mostly in seats that voted heavily to Leave the EU. Labour conversely saw huge changes in their share of the vote in seats that voted to Remain. Seats Labour held in 2010 appear not to be affected by the Remain/Leave divide. However, as the decade develops the seats Labour had control of starts to change and it appears Labour holds onto seats that would vote Remain in 2016, whilst throughout the decade they lost seats that would vote to Leave. As a result, Labour’s net losses to the Conservative Party comes through seats that voted to Leave, whilst nearly all their net gains culminate in seats that voted to Remain. This importantly indicates that the Remain/Leave divide has not always been present in UK constituency outcomes, rather it is a trend that has developed throughout the last decade, becoming stronger with each passing election. Therefore, as the EU issue rose up the agenda, and social divides deepened on this issue, so did the electoral divide social groups experienced. Areas that would go on to vote Leave drifted towards the Conservatives from 2010 onwards, whilst areas destined to vote Remain would increasingly back Labour post 2010.
There is further evidence to suggest that there has been a long-term trend across the Remain/ Leave divide. There has been a gradual decline in the majority sizes of both seats that voted to Remain and Leave, both for Labour and Conservative. Remain seats for the Conservatives have been declining in their “safeness” at each general election since 2010. Now most of the Conservative marginal constituencies are Remain seats. Labour is in 2nd place in most of these seats, but the Lib-Dems and the SNP are 2nd in some of these increasingly marginal seats to. On the other hand, Labour now are most at risk in Leave seats they just held onto during the 2019 election, with the Tories being best placed to win such areas. Therefore, the Remain/Leave divide appears to have been a large factor in dramatically reshaping constituency outcomes.
Across the decade the parties’ bases have changed in terms of their geographic location. Labour in particular has seen net gains in London and the south east, whilst receiving large net losses in the north. Alternatively, the Tories gained heavily over Labour in the north and Midlands region over a ten year period. Importantly, there appears to be a reversal in patterns of party support, with Labour doing much better in the south and the Conservatives being more reliant on the north. Delving further into the data it can be said that these trends are strengthened when focusing upon seats gained and lost by the two main parties across the decade. Seats that the Conservatives held in 2010, but lost control of by the end of the decade, showed similar regional patterns. In these seats there were large swings from Conservative to Labour, much larger than the average throughout the country. Moreover, the largest swings occurred in London and the South East, again indicating regions have realigned themselves between the two main parties throughout the last decade.
The seats Labour lost again displayed larger net gains for the Conservative Party than the average regional figure indicated was the case. The North East and West, along with Yorkshire and Midland areas to, displayed large swings. The areas Labour lost were traditional Labour areas and trends across the decade show large changes in the main parties’ share of the vote. The box plot underneath shows that the distribution of change per region is starker than the average figures suggested. For example, London shows that 75% of the seats has given Labour quite a large net gain, indicating a wide shift of support in Southern areas for Labour. The same can be said for northern areas for the Conservatives. For example, 75% of the constituencies in the North East gave the Conservatives a noticeable net gain over Labour, again indicating the realignment could be wider than regional average figures stated. The above evidence indicates that significant changes in geographical support for the main parties occurred, again show large shifts in the way British politics operates.
Going into more detail in geographical trends it can be stated that the geographical trends shaping changing patterns of party support are more detailed than has been widely discussed in media circles. Geographical trends have been summarised as a north/south divide where the loss of Labour’s northern red wall was the defining feature of the election. However, this ignored more detailed change that have occurred. Using a heat map to create a coloured scale that reflects the size of the increase, or decrease, in the percentage of the vote in every constituency for Labour and the Conservative across the last decade a distinct trend emerges, regardless of regional location. Seats that can be described as more Remain, urban, cosmopolitan, younger, highly educated and obtaining a workforce more reliant on professional occupations were less likely to decrease Labour support, and in some cases increased it. The red heat map below, representing changes in Labour’s share of the vote across the decade show Labour’s strongest gains occurred in specific constituencies. For example, Labour gained in London seats, Brighton, Canterbury, Liverpool and Manchester areas. The most interesting point to make here is that constituencies very close to areas where Labour gained, like Canterbury, Brighton and Manchester show the opposite trend where Labour received a decrease, represented by a light red colour. This indicates that rather than just being a general regional trend it is instead a constituency type trend. Labour did better in areas that have been more exposed to higher education, cosmopolitanism and globalisation, yet areas that have not been exposed to these factors relatively near-by have rejected Labour. Critically, areas representing places that are thought to be highly likely to reflect the culturally liberal side of the divide have grown in Labour support, whilst areas reflecting the culturally conservative side, regardless of their geographical location, have moved away from Labour and toward the Tories. Conversely, looking at the Conservative heat map the Conservative vote increased most in areas that can be described as Leave, older, less qualified, traditional working class areas. Oppositely, the Tories struggled to gain in the metropolitan areas Labour gained. Crucially, this indicates the Tories have lost support in areas most able to be described as culturally liberal, whilst they gained most in culturally conservative areas Labour lost in. This mirror reverse pattern shows that across the last decade Labour became more reliant on culturally liberal areas, whilst the Tories secured a new base in culturally conservative constituencies. This pattern occurred at a localised level and indicates a new divide around cultural issues, like Europe and immigration, are increasingly shaping where the two main parties can win and lose.
How influential are these trends upon changes in voting and election outcomes?
Table 4: This linear model estimates the effect certain social, political and economic factors have had upon the changes in constituency voting patterns between the 2010-'19 elections.
Model 1 displays how influential various factors were in changing the share of the vote the two main parties secured in a given constituency. It shows the Brexit divide to have been the most significant factor in shaping voting changes across the last decade, again showing the culturally liberal and conservative divide to have become increasingly important in shaping election outcomes as the decade progressed.
Variables representing the age divide were also found to be significant, along with qualification variables mostly being significant as well. This shows that younger people with more qualifications being more likely to vote for Labour as the decade progressed was unlikely due to chance. It also shows this was part of a longer-term process that was shifting political support in Britain. Conversely, the development where areas with older populations with less exposure to higher levels of education increasingly voting conservative over the decade also in all probability represents a long-term shift in how the electorate behaves.
Economic variables also have shown that the old left/right economic class divide has become less relevant in British politics. The regression analysis shows that class based voting patterns have moved away from traditional patterns. It highlights those in economic working class occupations shifted away from Labour and towards the Conservatives in large numbers across the decade at a highly statistically significant level. In later regression analysis it is found that this class divide no-longer limits Conservative support or bolsters Labour MP numbers. This indicates that the working-class trend moving away from Labour has made the old divides less redundant and hurt Labour’s ability to form winning coalitions, and from this win elections. Therefore, the class divide has spread across the two main parties fairly evenly as the decade progressed, making it less able to determine election results and give Labour a solid base of support. Moreover, the economic divide has weakened in other respects, where income deprivation and unemployment numbers have become unable to determine voting shifts from one election to another. Further to this, divides around housing occupation matter less, with those in socially rented sectors moving away from Labour and towards the Tories, again indicating class based voting has shifted more evenly across the two main parties.
Crucially, the regression analysis shows a trend, which is very unlikely due to random chance, where variables representing the culturally liberal/conservative divide have grown in significance and shaped voting outcomes at a greater extent than compared to economic factors across the decade. This caused votes to be split more evenly across the parties, and therefore economic factors became less able to explain election outcomes. Consequently, this all supports the theory that old patterns of support for Labour and Conservative have broken down. The new patterns of support, outlined earlier, are based around a new cultural divide that has occurred within British politics throughout the last decade, thus creating a new two-party system where the parties support bases have changed.
Table 5: This model estimates political, social and economic factors impact upon the probability of the Labour and Conservative Party to gain an MP.
The above regression model, model 2, outlines the probability of the Conservatives or Labour winning a constituency in 2019 based on various variables in the constituency dataset. The model demonstrates that the percentage of voters within a constituency that voted to leave the EU was very significant in shaping the ability the two main parties had in securing an MP, and from this winning the election. The effect from this model shows that the larger the majority for Leave was within a constituency the more likely the Conservatives became in winning a seat, with Labour’s chances declining. Conversely, Labour became more likely to retain a constituency when a constituency voted to Remain, and in some cases this actually increased Labour support. Significantly, this shows that divisions from Brexit may have carried on from 2016 and translated into different levels of support for the two main parties. These changes became so great that in all probability this was a large contributing factor in why the Conservatives were able to gain enough seats to win a large majority.
Further to this, economic variables representing the economic workforce, broadly a measure for different class groups, were found to be insignificant in determining the election result. This would indicate that whilst some studies found class to still have some significance in the 2010 and 2015 elections at a constituency level this has completely been eroded by 2019. This is hardly surprising as the constituencies Labour lost clearly were disproportionately working class and therefore it is not surprising to find that class loyalties no longer are strong enough to give any party an increased probability of representing an area. However, age and qualification variables were mostly found to be significant in this election, a change from past elections. Therefore, whilst economic divides have become less significant social and geographic ones appear to have become more useful in explaining election outcomes.
What does this all mean, the Bigger picture:
In terms of understanding British politics during the 2010s this article encourages students and researchers to analyse political developments through the developments set out within this blog. When understanding such developments within British politics this should be thought of as a gradual change that occurred slowly across the decade through successive elections. Labour lost in Leave areas in successive elections throughout the decade and the leave variable is a clear factor that can explain changes from their 2010 vote share compared to their result in the 2019 election. However, this should not be oversimplified as just a Leave vs Remain divide as many Labour members would wish to believe, instead it should be viewed as a much wider and deeper change that represents a shift in the party’s base of support. Labour did not just lose in areas that heavily voted to leave, they also lost out in areas that voted to leave by a small margin, but also shared characteristic traits with seats that Labour no longer control in red wall areas. Instead, rather than just a Leave/Remain divide electoral change during the 2010s should be thought of as a wider social change where older, de-industrialised working-class towns away from university and cosmopolitan influences have shifted away from Labour and bolstered Conservative support.
This has altered party politics as now the Conservatives should be thought of as a party that had changed from Cameron’s Conservatives, who tried to secure middle-class liberals in university areas, to a different more culturally Conservative Party which aims to win over working class areas under Johnson’s leadership. Meanwhile, Labour can now be thought of as the party of culturally liberal constituencies, even ones thought of as once safe Conservative seats. Therefore, party competition in this time-period must be understood as one of great change that was re-shaped by the culturally liberal/conservative divide that had formed over a long-time due to wider economic and social factors, such as globalisation.
The future for British politics?
The swing towards the Conservatives in these culturally conservative areas across the decade has been so large that these seats are unlikely to revert back to safe Labour seats anytime soon. Moreover, Labour’s gained in some areas so much in areas they now represent t they can be thought of as mostly safe seats, such as Brighton. Therefore, these large changes in support are unlikely to be reversed anytime soon and British politics, and its electoral competition, therefore will continued to be focused around such areas.
Therefore, Labour will find it tough to win the next election as they may struggle to represent both the winners and losers of globalisation. Indeed, Labour could lose more seats that nearly fell to the Tories in 2019 if they fail to understand and represent this new culturally liberal/conservative divide.
Finally, this research critically supports Curtice's research which has found a realignment may have increased the likelihood of a hung-parliament election outcome as fewer constituencies mirror these divides enough to produce enough constituencies that can change hands to ensure an overall majority outcome. Therefore, even if Labour does claw back support, due to the long-term loss of working class areas and Scotland a hung-parliament could be the most likely scenario, possibly making more unstable political times more likely in the near future.
Note: If you would like to request the methodology behind this study please email the author James Prentice, James.Prentice@live.co.uk. Also, if you would like to request this chapter of my upcoming thesis in full please email me at the stated email address.