It’s more complicated than that!: Unpacking ‘Left Behind Britain’ and some other spatial tropes following the UK’s 2016 EU referendum

Introduction

In a letter published in the Northern Echo newspaper in April 2019, four UK Labour Party Members of Parliament (MPs) called for the ‘stereotyping and the caricaturing’ of the north of England in the wake of the UK’s 2016 EU referendum as ‘Brexitland’ to stop (Creagh et al., 2019). Referring to Winston Churchill’s dictum that ‘There is no such thing as public opinion. There is only published opinion’, the four MPs drew attention to how ‘published opinion’ has crudely represented the north of England as the fulcrum of support for the ‘Brexit’ project, arguing that ‘Journalist after journalist writes it. Broadcaster after broadcaster says it out loud. The London-based metropolitan media has declared it so. The North is Brexitland’ (Creagh et al., 2019). The letter thus alludes to the problematic nature of the dominant view which has arisen since the UK’s 2016 referendum on membership of the European Union (‘the referendum’), which sees the result primarily as an outcome of the uneven geographical changes produced from the economic restructuring of the latter decades of the 20th century. The latter, it is argued, has benefitted certain regions, notably London and the South East and some larger centres elsewhere, but often bypassed many other places such as smaller industrial towns and cities and deprived coastal areas (Burrell et al., 2018). The argument is that the distribution of the vote to leave the EU reflects a revolt of such places – a view often associated with the invocation of an emotive and arresting spatial imaginary of ‘Left Behind Britain’ (LBB) (Sykes, 2018). This spatial-narrative trope has been associated in many media analyses with the wider populist ‘moment’ being experienced in different parts of the developed western world in the mid-to-late 2010s (Eatwell and Goodwin, 2018). It also seems to provide an archetypal instance of a spatial imaginary that has achieved powerful ‘performative agency’ (Watkins, 2015), in shaping and in many cases jumbling thinking about, and responses to, the referendum result.

Informed by this context, we seek to unpack and problematise some of the assumptions about the geographies of the referendum result and to nuance and highlight certain contradictions in the hegemonic ‘LBB’ spatial imaginary of its causation. We start by considering what we know about the national regional geography of the referendum result, before discussing some of the most commonly discussed spatial imaginaries, or tropes, which have come to characterise the referendum vote. Following this we use detailed voting data from one of England’s core city regions to consider the complex relationships between voting and socio-economic geographies as a means to reflect on how (and where) such tropes do provide satisfactory explanations and where they do not.

The ‘short handing’ of the geographies of the EU referendum

Manley et al. (2017) observe how ahead of the 2016 referendum most analysis was ‘based on opinion polling which focused on which groups were more likely to support each of the two options, with less attention given to the geography of that support’ (183). In contrast, since the referendum this situation has arguably been reversed with geographical explanations of the result becoming extremely prominent in the ‘published opinion’ of media and academic commentaries. Many geographical representations of this are crude, presenting simple visualisations of Leave/Remain areas which hint at a geography where ‘Remain’ votes are predominately cloistered in the largest English cities, and Scotland, and ‘Leave’ sentiment permeates much of the rest of the land. Yet, in line with Manley et al. (2017), we argue that at the UK ‘national’ scale, the voting patterns of June 2016 have frequently been presented and discussed in ways which have oversimplified the geographies of the result. In many cases, this oversimplification has been accompanied by imaginaries and caricatures of ‘leave’ and ‘remain’ areas and forced ecological fallacies in the presentation and interpretation of the data. Although some mappings present data at a local authority level (the scale at which referendum results were initially reported), some mappings and analyses are even starker, with results simply aggregated to whole nations of the UK (England, Wales, Scotland, Northern Ireland).

Sometimes, these visualisations have then been interpreted using the same ‘first past the post’ logic as used in Westminster elections, with talk of how many areas voted ‘leave’ or ‘remain’. Yet, given that the purpose of the ballot was not to send representatives to Parliament but to answer a question at the UK scale this is problematic – not least as it creates an impression of a far more one-sided result than actually occurred. Indeed, if we move away from the binary outcome of the vote, but rather consider the difference between the two outcomes in an area, the overall image becomes less stark. For example, there were 123 authorities where the difference between outcomes was less than 10% (Figure 1), i.e. the point at which an area might ordinarily be considered a battleground or marginal. Taking this further, whilst there were 82 places where the vote outcome (either to Leave or Remain) exceeded more than 20,000 votes (35 remain, 47 leave), in 16 authorities the difference was less than 1000 voters.

OPEN IN VIEWER

Figure 1.

Local authorities considered ‘battlegrounds’. Source: Electoral Commission (2016).

Analysis on first past the post terms also often stumbles against its own logic, obfuscating the fact that in the UK’s quasi-federal union of different nations half the constituent territories voted to leave and half to remain, and in the intervening period subtly and politically different ‘aggregations’ and scalings of the results have been used to support different positions. The powerful sentiment in some of the devolved territories, for example, has been that the UK government has sought to monopolise the management (such as this has occurred) of the response to the referendum result. This sentiment of marginalisation is, perhaps unsurprisingly, also acutely felt in the UK’s largest territory, England, where, for example, city regions with populations that exceed those of a number of EU member states have barely been consulted on the form that the process now known as ‘Brexit’ from the EU may take (Rankin, 2019).

In attempting to make sense of the referendum result, the broader discussion has been heavily coloured by other contemporaneous issues and value-based explanations (Kaufmann, 2016) which only serve to complexify territorial readings of Britain’s current predicament. The most pervasive of these explanations is the presence of what has often been characterised as a ‘Left Behind Britain’ (LBB) – a group of people who feel disconnected from the broader success of the UK. Yet this LBB can present itself in many forms. It can be characterised by communities which have been particularly affected by economic shifts caused by globalisation (Sturzaker and Nurse, forthcoming), have seen their economic performance stagnate (McCann, 2016), and in which many people lack the qualifications (Manley et al., 2017) to access new economic opportunities, e.g. in emerging sectors.

Another common reading within the same narrative is that LBB and the ‘Brexit’ vote is a function of a decade of regressive austerity which has affected local areas – particularly in the northern parts of England (Centre for Cities, 2019). The geographical effects of this austerity are, ironically, laid bare in figures relating to the next round of EU Cohesion Policy (2021–2027), in which some areas – including many which voted to leave – could see their EU regional funding allocation increase by around 22% (CPMR, 2019). This increase in funding eligibility is a reflection of the stark reality that areas of the UK are, as in the 1980s, falling behind the EU average in terms of prosperity (Centre for Labour and Social Studies, 2018). This is problematic policy-wise, not least in the face of the nationally proposed replacement initiatives for EU Cohesion funding such as the UK Shared Prosperity Fund (Brien, 2019). The latter is still in its gestational stages with a paucity of information in the public domain about the timing of its launch and its quantum. If eventually launched it will face scrutiny in terms of how it measures up to the potentially foregone multi-annual support from EU structural and investment funds. The political salience of the LBB narrative is also reflected in the ‘Stronger Towns’ fund, first announced in March and later rebranded as simply the ‘Towns Fund’ (MHCLG, 2019), where it has been calculated that 94 of the 100 towns slated to receive funds from this voted to Leave (Williams, 2019).

The focus on ‘Left Behind Britain’ is only one reading of the territoriality of the referendum result, yet as Sykes (2018) and Burrell et al. (2018) have argued the LBB narrative has become the dominant reading of the geography of the referendum and gone on to exert a powerful influence over perceptions. Reporters and TV film crews have beaten a path to places like Sunderland, Stoke on Trent, and Boston (Lincs.) to gather almost anthropological ‘vox pop’ style reports which reinforce this LBB narrative and which, in the words of the Labour MPs mentioned in the introduction, depict:

‘“the north” with films of chimney pots and run-down mills, [and] angry men who warn that if Brexit is not delivered, immediately, there will be riots on the streets’ (Creagh et al., 2019).

Hence, the LBB spatio-social imaginary, or trope, does useful work for the leave lobby conditioning public opinion to see the choice to leave as being the will of this ‘Left Behind Britain’. Yet, this overall spatial characterisation has already been challenged (Dorling, 2016; Dorling and Tomlinson, 2019) with Dorling (2016) pointing out that although some such regions did vote to leave the EU, this was against a wider backdrop where:

Contrary to popular belief, 52% of people who voted Leave in the EU referendum lived in the southern half of England, and 59% were in the middle classes, while the proportion of Leave voters in the lowest two social classes was just 24%.

It is these views which feed into another commonly held analysis which gained particular traction: a young versus old dynamic. Again, there is some underlying evidence to support this view, not least a spate of opinion/exit polls taken in the period near the referendum (Curtice, 2018), alongside post-result analysis (Manley et al., 2017) which suggests that younger voters (those under 45) were more likely to vote Remain, whereas older voters (those over 45) were more likely to vote Leave.

This has led to the emergence of a further key trope around the nature of the vote: that older generations were, effectively selfishly, voting to leave the EU against the expressed wishes and interests of their children and grandchildren. This fed on broader tensions around inter-generational fairness which had been brewing for some time, and particularly began accelerating in the wake of the 2008 Financial Crisis (Bristow, 2017). In this trope, having benefited from a society in which affordable housing, free higher education, a ‘job for life’ followed by generous pension that is accessible at a modest (or early) retirement age were all readily available, the so-called Baby-Boomers were voting against the interests of a younger generation who had no-little prospect of these things and, in doing so, were actively ignoring warnings of economic harm which may damage those prospects even further (Elledge, 2017). Ultimately though, this demographic trope also represents an oversimplification of an issue that will be explored further below.

Another theme of discussion surrounding the EU referendum was the voting patterns amongst ethnic minority voters. The latter had been somewhat crudely wooed by the ‘leave’ campaign with promises such as more work permits for those from non-EU countries, ignoring the fact that international immigration policy is a national and not EU-level competence (Haque, 2018). This was a revisiting of the electoral strategy of division, seeking to play different groups, notably different immigrant communities, off against one another, cultivating a sense of inter-group grievance, invoking notions of ‘last one in, first one out’, and offering clientelistic promises of future benefits in return for electoral support. Yet in the event the nostalgia for empire identified by Dorling and Tomlinson (2019) which worked so well in other quarters seemed to have a dragging affect with ethnic minority voters, with Begum (2018) arguing that ‘the xenophobic and anti-immigrant tone of the Leave campaigns’ and

Slogans such as ‘take back control’ and ‘make Britain great again’ were less appealing to ethnic minority Remain voters, who associated them with nostalgia for empire and a longing for a ‘pre-immigration white era’ on the part of the Leave campaign.

As we can see, in the aftermath of the referendum, there emerged a number of persuasive, and often pervasive, analyses to try and make sense of the vote. Informed by this context, the rest of this discussion attempts to unpack some of the issues raised by these representations of the results of the EU referendum. Prefaced by a brief national-level analysis of the voting data which examines the broader geographical presentation of voting patterns, the main thrust of our analysis is informed by ward-level data from two local authority districts that voted to remain within a wider city region (Liverpool City Region (LCR)) which voted remain overall, but where half the local authority districts voted to remain and half to leave. It presents data from wards that have similar socio-economic conditions (e.g. with similar ranks in the Index of Multiple Deprivation (IMD)), but which voted differently in the referendum.

Case study and method

Our understanding of the various geographies of the referendum has been compounded by the fact that in most cases data on voting patterns have largely only been available at the local authority level and, beneath this, there is a stark lack of granularity. This lack of data, we were told by one local councillor, reflected an instruction from central government to report the results as quickly as possible, with little focus on the detail – the result being that on election night the results were collated and reported at a local authority level. At the slightly more granular scale, voting data for parliamentary constituencies does exist – something used by various pressure groups in their attempts to motivate particular MPs with regards to voting on aspects of the subsequent ‘Brexit’ process. However, these data can be premised on estimates – not least where larger local authorities such as cities often comprise multiple constituencies. For the most part, this represents the limits to the granularity of publically available voting data and, in contrast to other elections, there is scarce detail at the ward level – in most cases because it simply did not exist. Thus, for the most part, and as we have discussed above, the various tropes which have gained traction as people attempt to rationalise and understand the drivers of the referendum result are based on a mixture of high-level voting data, opinion polling, and piecemeal accounts.

Amidst this overall sparseness, however, a small number of local authorities did collate data at the ward level and, more importantly, made it available. Some did this immediately after the referendum; others released it after some time. This paper centres on two local authority districts where this occurred: Liverpool and the Wirral. That these two authorities are amongst the few areas where ward-level data on the referendum exist provides an initial rationale for their selection, but they are useful in other ways.

First, they are contiguous parts of the same city region: LCR. In this, they share similar characteristics, both in terms of post-industrial economic landscape – both have been profoundly affected by the decline in dock/port-related activity (Sykes et al., 2013; Wilks-Heeg, 2003) – strategic governance (Sturzaker and Nurse, 2019), and socio-economic profile. Consequently, they should be emblematic of at least one characterisation of ‘Left Behind Britain’. Second, the wider political context of the two authority areas is important: that they both returned Labour MPs to all their constituencies at the last general election (though two of these have subsequently ceased to represent the party) makes them valid locations in which to examine some of the tropes which been used to explain the referendum result. Third, both authorities voted remain in the referendum and are within a city region that (if votes are aggregated at this level) voted to remain, but where half the districts voted to leave (Table 1).

OPEN IN VIEWER

Table 1.

Voting data for the Liverpool city region.

Local authority	Voted remain	Voted leave	% Remain	% Leave
Halton	27,678	37,327	42.6	57.4
Knowsley	34,345	36,558	48.4	51.6
Liverpool	118,453	85,101	58.2	41.8
Sefton	76,702	71,176	51.9	48.1
St Helens	39,322	54,357	41.0	59.0
Wirral	88,931	83,069	51.7	48.3
Total	385,431	367,588	51.2	48.8

Source: Electoral Commission (2016).

The data presented within these areas offer a useful opportunity to explore the extent to which those tropes do hold up at the micro-level. In doing so, our aim is to give further consideration to those explanations which might have traction and those which might not. This is important – not least for policy makers trying to provide meaningful responses at the local and national scale.

To achieve this, we consider a mixture of relevant contemporaneous data and assess this against the granular spatial voting patterns. Where possible, and to provide the greatest level of context, we consider this data at the Lower Super Output Area (LSOA) level – effectively neighbourhoods of approximately 1500 people. In selecting our data, and line with the discussion above, we explore the extent to which ‘leave’ and ‘remain’ sentiment aligns with three themes – each reflecting the dominant tropes surrounding the referendum result discussed previously. They are as follows:

‘Left Behind Britain’ – To consider the commonly asserted view that the Leave vote was driven by those who feel disconnected from the broader socio-economic prosperity of the UK, we use the IMD – effectively a measure of social and economic deprivation which ranks more than 32,000 LSOAs in England. Conducted on a semi-regular basis, the most recent IMD data were released in 2015 (DCLG, 2015), less than a year before the referendum, providing an accurate picture of where those disconnected areas would have been placed at the time of the referendum. In this paper, we consider IMD ranking by ‘decile’, i.e. if a neighbourhood is in the bottom 10% of the national ranking, and so on.

As we will see, the data reinforce the arguments of those who have claimed that the phenomenon of Brexit is powerfully cultural and contextual and that socio-economic analyses of its causes do not fully explain why some areas and populations voted to leave the EU whilst others with comparable profiles voted to remain. With poorer regions predicted to be the biggest economic losers of ‘Brexit’ (McCann, 2016, 2018) our understanding of such issues is of material consequence and, regardless of the ultimate outcome, may give clues in formulating responses in its aftermath.

Maps and analysis

As we can see in Figure 2, which maps the referendum result for Liverpool and the Wirral, the presence of ward-level data allows us to move from a blanket result towards a more spatially distributed visualisation of voting patterns. Immediately, it is clear that in both authorities there are clear spatial variations in how people voted. In Liverpool, the Remain vote is concentrated in the city’s core, its immediate surrounds, and towards the south. In the Wirral, there is a clear East–West divide. Already, this suggests that we can dispense with one trope, in that the top-level result from a local authority does not mean that the electorate voted as a bloc in an un-nuanced way, but rather that there are, as might expected, significant pockets of ‘opposing’ sentiment in the face of the returned overall result for a local authority area.

OPEN IN VIEWER

Figure 2.

EU Referendum Voting – Wirral and Liverpool. Source: Wirral Globe (2016) and correspondence with Labour Councillor (2018). https://www.wirralglobe.co.uk/news/14582313.detailed-breakdown-of-how-wirral-voted-in-referendum-the-poorer-the-area-the-bigger-its-leave-vote/

Acknowledging that there are likely to be pockets of support for both outcomes within any area is important – and in this way directly mirrors the problems with current national-level analysis of the Referendum result. Whilst broader democratic principles dictate there must be a ‘winner’, a broad brush representation of the results can be dangerous, not least within a representative democracy, where glossing over the nuances of a result might pave the way for more zealous voices (of many persuasions) to overlook the views of their constituents. Side-lining significant proportions of the electorate in a major national decision is a strategy that risks undermining the overall political legitimacy of governing decisions and undermining the search for consensus on divisive issues.

‘Left Behind Britain’

When considering deprivation in both Liverpool and the Wirral (Figure 3), the evidence to support idea that the Leave vote was driven by a ‘Left Behind Britain’ is mixed. In the Wirral, where deprivation is concentrated in the urban centre of Birkenhead, the Leave vote does broadly align with this spatial concentration, whilst the Remain vote is at its strongest in the least-deprived western wards. Here, there is evidence which lends credence to the idea that a ‘Left Behind Britain' is a key driver of the ‘Brexit' vote.

OPEN IN VIEWER

Figure 3.

Index of Multiple Deprivation (2015) by decile. Source: DCLG (2015).

Liverpool, however, presents a very different picture. With roughly half of the city located in the bottom decile (i.e. the 10% most deprived areas in the country), and as the site of some of the single most deprived neighbourhoods nationally, there is no doubt that those areas embody the idea of ‘Left Behind Britain’ – bearing the hallmark scars of post-industrial realignment (Sykes et al., 2013) and, most recently, the austerity which has decimated public spending (Centre for Cities, 2019; Lowndes and Gardner, 2016). Yet, in large part, and contrary to the what might be expected from the ‘Left Behind Britain' trope, the significant majority of those areas voted to Remain and, going further, some places such as Toxteth (a deprived ward to the South of the city centre), had the proportionally highest Remain vote. Similarly, the four wards which voted to Leave present a mixed image – not least in the North East of the city where the Leave vote is underpinned by some of least deprived LSOAs.

It is important to acknowledge that there are some areas of Liverpool which do align with the ‘Left Behind Britain' trope (i.e. greater or lesser deprivation aligns with a proclivity to vote leave or remain, respectively), but the overall image of the city is not one where voters neatly align with this pattern. In effect, ‘Left Behind Britain' does not describe the city’s voting choices in a convincing way. It encapsulates some areas, but not others even though they share markedly similar socio-economic profiles.

It is not unreasonable to ask why this might be. In the lead-up and aftermath of the Referendum, much was made of ‘Metropolitan Elites’ – used to pejoratively describe city-dwelling white-collar workers considered to be out-of-touch with the national mood (i.e. anti-EU sentiment). But it would be patently wrong, to suggest that merely living in a city makes you ‘elite’ – something supported by the UK’s urban history from the Industrial Revolution onwards (Couch, 2016) and clearly observable in Figure 3. But going further, in the case of Liverpool it is also evident the ‘metropolitan non-elites’ (to use the same phraseology) voted in a broadly similar manner, and thus this provides a similarly unsatisfying explanation.

Ultimately, the ward-level data for Wirral and Liverpool give a mixed verdict on the LBB narrative. There are examples where it holds up, particularly in the Wirral, but others where it falls down to the point where the reality runs entirely contrary to expectations. Parts of this area are the very embodiment of a ‘Left Behind Britain’ but they do not act in the way that dominant representations of the territoriality of the vote would suggest.

Age

To assess how age aligns with voting patterns, Figure 4 illustrates the percentage of population in each LSOA categorised as 45 or above in the 2011 Census. Within these data there are two broad patterns that we can observe with regards to the age distribution in Liverpool and the Wirral.

OPEN IN VIEWER

Figure 4.

Percentage of population aged >45 at the 2011 Census.

The first is that there is a significant cluster of younger people (i.e. LSOAs where more than 80% of the population is aged under 45) which protrudes south-eastwards from Liverpool city centre. This can be attributed to two elements: young professionals living in the city centre, and Liverpool’s 50,000 strong student community who can also be found in the city centre and the Smithdown Road corridor to the South East (Mulhearn and Franco, 2018). Reflecting the transient but churning nature of the student community, whilst those students present in 2011 are highly likely to have moved on after completing their studies, we would suggest that such clustering remains accurate. The second is that the Western part of the Wirral, along with a pocket in the north east coast of the peninsula, is dominated by those who are over 50.

As with ‘Left Behind Britain', a straight analysis of Old V Young provides a similarly unsatisfactory fit in describing the voting patterns of Liverpool and the Wirral – in some cases even less so. As above, there are instances where this interpretation is a good fit, not least when describing the younger population in Liverpool – where those in the city centre and student areas did vote to Remain. However, those student areas only partially overlap with the most heavily remain-voting areas to the South of the city centre.

When considering older voters, however there is far less of a clear fit with this narrative. For example, in Liverpool, the areas with older populations (Woolton and Calderstones) voted Remain, whilst in the North of the city there is no clear pattern. Similarly, in the Wirral it is those areas of older population (i.e. west and North West Wirral) which voted heavily to Remain vote and those areas of younger overall population that correspond to the Leave vote.

Taking this further, when these data are taken in line with the previous discussion on deprivation and LBB, our findings recall those of Manley et al. (2017) in that, at the very least, older voters from less-deprived areas appear to show a greater proclivity to vote remain. This suggests that the idea of older versus younger voters can be a crude dichotomy to adopt at a more refined spatial scale, although there remain further questions as to what separates the older, wealthier voters of Liverpool and the Wirral with their Eurosceptic equivalents in Southern England identified by Dorling (2016) and Dorling and Tomlinson (2019).

Ethnicity

Figure 5 shows ethnicity according to the 2011 Census, charting the percentage of population identifying as ‘White’. As we can see, although Liverpool and the Wirral are majority white areas, there are some exceptions to this. In particular, and reflecting its history as a major port in the British Empire (Sykes et al., 2013; Wilks-Heeg, 2003), Liverpool is home to two large non-white communities. The first is Liverpool’s Chinese community, which originally concentrated in the city’s Chinatown area that abuts the city centre area, and contains the largest Chinese arch outside of China. The second is Liverpool’s black community located primarily in the ‘L8’/Toxteth area to the immediate south of the city centre – largely composed of residents with a Caribbean and Somali heritage.

OPEN IN VIEWER

Figure 5.

Percentage of population identifying as ‘White’. Source: NOMIS (2011).

In comparing Figures 2 and 5, it is clear that areas with different proportions of population identifying as white voted to Remain. Beyond this, ethnic composition of areas seems weakly related to how any other area voted – with some places with the highest proportion of population identifying as white voting to Leave, and others to Remain. The Princes Park ward south west of the centre of Liverpool with a 51.2% BAME population (Liverpool City Council, 2018) voted 73% to Remain, however, tending to confirm the national tendency for ethnic minority voters to strongly support remaining in the EU.

Conclusion

This discussion has sought to consider some of the socio-spatial tropes and caricatures that have come to characterise the aftermath of the UK’s 2016 EU referendum, and argued that some of these rest on insufficiently granular data and evaluations of the result. At their crudest, depictions which aggregate the result to the four nation territories of the UK, or attribute local authority or constituency areas to ‘leave’ or ‘remain’ in blanket fashion based on simple majoritarian logics, shorthand the finer geographies at play, blunting our understandings and thus our ability to respond appropriately.

It is clear too that some readings of the spatiality of the vote have acquired greater dominance than others in understandings and responses in the aftermath of the referendum and in ‘published opinion’. The most pervasive of these has been the social-spatial narrative which describes the revolt of a deprived northern and peripheral ‘Left Behind Britain' or ‘Brexitland'. Indeed, and despite sitting uneasily against the early evidence which suggests that in absolute terms the key territories of the leave result were to be found in areas of affluent Euroscepticism primarily in southern England, this narrative has exerted a powerful influence on perceptions. The aggregate data illustrating that older voters had a greater propensity to vote leave have also animated debates with the referendum result being interpreted in terms of intergenerational equity and tensions. The interests and votes of BAME voters have largely been ignored in the wider post-referendum debates, and even the spike in hate crimes which followed has been largely viewed as a collateral issue, with little introspection noticeable amongst those who employed anti-immigrant rhetoric and imagery in the 2016 campaign.

The present paper has sought to explore some of these themes in the context of two local authority areas in a northern English city region for which voting data from the EU referendum are available at ward level. It has set this alongside data relating to deprivation, age profile, and ethnicity to offer some reflections on where given macro spatial-social tropes which have risen to prominence since 2016 hold true and where they appear to be challenged by the descriptive statistical data. The analysis cannot claim to be comprehensive and a fuller picture would require further work to ascertain contextual and causal factors and links to attitudes in different areas. But the data reviewed does show where some of the wider assumptions and socio-spatial narratives which have been used to frame the ‘Brexit’ debate since 2016 appear to be confirmed, or begin to breakdown, when finer geographies are considered.

Thus, within the two local authority areas, one can find both relatively affluent and more deprived areas that voted strongly to remain, but with a contrast between Liverpool where relative disadvantage seems less linked to a ‘remain’ or ‘leave’ result and the Wirral where (in-keeping with the LBB trope) more affluent areas tended to vote remain and less advantaged areas tended to vote leave. In the words of one local councillor we spoke to ‘The left behind Britain white working class arguments don’t play in Liverpool’. Similarly, and in line with Manley et al. (2017), the ‘ageist’ interpretation of the referendum result is rather challenged by the fact that some areas with relatively older population in both areas had strong remain results. Meanwhile, a younger overall population profile was associated with a strong remain result in central and east-of-centre areas of Liverpool with their large populations of younger adults, but not in areas of the Wirral with a younger age profile which tended to vote leave. In terms of BAME voting patterns, the results tended to confirm the wider national picture in that areas with concentrations of such voters voted heavily to remain, despite the leave campaign’s attempts to offer inducements to some groups and ‘divide and rule’ between more established and more recent immigrant communities.

Overall, the findings hint at the inadequacy of crude material socio-economic analyses of why individuals voted the way they did, e.g. why did areas with similarly high levels of deprivation in Liverpool and Wirral vote so differently? Or at a higher city regional level why did local authority areas with similar IMD profiles vote so differently? Why did relatively deprived areas with a population predominantly identifying as white vote remain in Liverpool, but leave in Wirral? Questions such as these direct attention to the need for greater contextual understandings and appreciation of cultural and place-based factors. Such understandings will become important if the (present) proto-‘Brexit’ state, or any future post-Brexit (leave or remain) state ever decides to pick up the mantle of developing a credible place-based policy to address the impacts of the socio-economic change of the austerity decade of the 2010s. Finally, this paper was written at a moment when there is an increasing (belated?) awareness that some of the dominant narratives and ‘published opinion’ which have accumulated around the EU referendum is in many cases too crude and oversimplified, and from the perspective of certain political agendas potentially counterproductive (Creagh et al., 2019). This is perhaps exemplified by the heckling of Prime Minister Boris Johnson in mid-September 2019 in walkabouts and staged events across Northern England – the heart of this supposed ‘Brexitland’ (Woodcock, 2019). Similarly, it is noteworthy that the earlier results of the local elections in England in May 2019 posted losses of council seats for the Labour Party, and gains for the unambiguously ‘pro-remain’ EU Liberal Democrats and Green Party, in areas which have been caricatured as belonging to LBB (notably in North East England). In light of such developments, more work is clearly needed to develop the reflections we have offered here, but based on these and a growing body of evidence (Dorling, 2016; Dorling and Tomlinson, 2019; Manley et al., 2017), our response to the argument that the vote to leave the EU reflected a revolt of ‘Left Behind Britain’ would simply be that ‘It’s more complicated than that!’.