Brexit and the Politics of Housing in Britain

Abstract

political earthquakes—both real and perceived—are trembling through Britain. In the 2015 general election, the Scottish National party and the UK Independence party mounted successful challenges from the periphery. In the 2016 referendum on EU membership, the Vote Leave campaign captured a surprising number of votes from both sides of the traditional political divide. And in the 2017 general election, an unlikely Labour leader turned metropolitan areas red, while the incumbent Conservative Prime Minister turned post-industrial towns blue—bringing the whole party system back into a familiar bipolarity. At this breakneck pace of political change, scholars have struggled to make sense of what is driving British politics, and how. In this paper, we argue that the answer to this question might lie right under our feet. Over the last three decades, the British housing market has undergone a rapid transformation. In the postwar era, British housing featured high levels of social housing and low levels of housing wealth inequality. Today—following a mass social housing sell-off and a historic boom in asset prices—it features the inverse. We argue that this transformation has had a profound impact on British politics by dividing regions, tenures, and generations in a new housing cleavage. To make this case, we examine the recent history of British housing and its relationship to political changes both past and future. We begin by arguing that the performance of the British housing market—both the cause and the effect of local economic fortunes—shaped attitudes toward Britain's relationship to the European Union by structuring voters’ sense of inclusion in the project of international integration. We argue that house prices, even at a highly disaggregated local level, strongly predicted the vote for Leave or Remain. We then turn to the post-Brexit period. Our analysis suggests that housing played a similarly crucial role in the 2017 general election, with parties divided over their representation of different tenures. We conclude with a discussion of the prospects for housing policy reform in the context of new tenure coalitions. Real estate has come centre stage in the story of advanced capitalism. While advanced industrial economies relied on manufacturing to drive economic growth in the postwar period, today they rely in large part on the performance of assets like real estate.1 Citizens who once relied on wages for income and pensions for retirement now rely on credit and real estate.2 But Britain presents an extreme case. Since 1971, the stock value of British real estate has risen from roughly $60 billion to over $6 trillion—a 100-fold increase, adjusted for inflation. Housing now accounts for more than 44 per cent of all household wealth.3 The figures are equally striking in terms of real estate's flow contribution to the British economy. Real estate accounts for more than 12 per cent of Britain's gross domestic product—larger than the entire manufacturing sector4 —the majority of which comes from house prices through ‘imputed rent’: hypothetical rental income from owner-occupied homes. The performance of the British economy has long been closely linked with its housing market and, as we shall see, so too is its politics. Figure 1 uses data on (nominal) mean house prices at the ward level (around 5,000 people) between 1996 and 2016. For each year, we order the wards by their mean house price and show the overall distribution across wards, with the circle reflecting the median ward. Two things jump out. First, average nominal prices rose dramatically across the period—from £60,000 in 1996 to £250,000 in 2016, a quadrupling of nominal prices during a period where UK inflation rose just 76 per cent. Second, the distribution of average prices across wards also rose dramatically. The gap in prices between the wards at the 75th and 25th percentile rose from around £35,000 in 1996 to around £185,000 in 2016. Property in Britain has become not only more expensive relative to other goods, but also much more unequally distributed across the country. The broader economic effects of this transition have also been highly unequal. For those who own property in areas like the south of England, rising house prices have heralded greater prosperity. According to Andy Haldane, the chief economist of the Bank of England, property today is a ‘better bet’ for retirement than a pension.5 For those who rent property, however, rising house prices have heralded greater exclusion. In 2015, the average private renter in Britain paid more than 28 per cent of her income on rent. In London, that figure rose to 40 per cent.6 Housing markets also engender inequality in the daily lives of owners and renters. While homeowners enjoy security of tenure, renters in Britain have very few protections compared to their European neighbours. Section 21 of the 1988 Rent Act permits landlords to evict tenants with two months’ notice without citing a grievance. Section 21 evictions have spiked alongside rising property values, as many landlords pursue new tenants to pay higher levels of rent. The result is increasing inequalities of space. While homeowners have enjoyed an increasing ratio of rooms per person, renters have been pushed into smaller units because that is what they can afford.7 The transformation of the British housing market is both product and producer of a new geography of fortune. Regions like the south of England have been well served by the service sector transition of the British economy, while the traditional manufacturing powerhouses in the north of England have suffered. The former have enjoyed higher levels of investment and the inflow of workers that accompany it, driving up demand for housing. The latter have lost out, and the housing market reflects this dearth of demand. The geography of house prices tells the story of the British wealth and well-being. The enormous disparities seen above reflect, in part, the longstanding differences between Britain's regional economies since industrialisation. It is no surprise that homeownership has historically correlated well with Conservative party support:8 richer, more Conservative-supporting areas, tended to have higher house prices on average. The housing boom and bust of the past two decades has to some extent accentuated these divisions. Yet they have also unearthed new political divisions that cross the traditional geographic patterns of voting. In particular, the recent ‘triumph of the city’—wherein inner cities moved from apparently crime-ridden and destitute areas to become highly desirable, well-positioned neighborhoods9 —has created an unfamiliar political geography in Britain. The Brexit referendum of 2016 exposed these new cleavages. The Leave/Remain patterning of vote across local authorities in England and Wales did not match standard Labour/Conservative patterns. In particular, a large number of wealthy cities, towns and nearby rural areas in the south-east and south-west voted Remain, despite a long history of voting Conservative. Similarly, many poorer towns in the Midlands and north-east, traditional Labour strongholds, voted to leave the European Union. Split in this way, both parties remain fundamentally divided on how to approach Brexit. What does explain the geographical pattern of voting in the EU referendum? Ansell argues that house price levels and changes at the local authority level are strong predictors of support for remaining in the European Union.10 In particular, he shows that local authorities with houses twice as expensive had support levels for Remain around 15 percentage points higher, even controlling for their region in the country and local demographics such as age profile, migration rates, unemployment, and local wages. Although data were only officially collected at the local authority level, the BBC was able to acquire ward voting rates for some 1,100 wards. We analyse this data to see if the pattern of high house prices and Remain voting holds at a very local level. Table 1 shows results from statistical regressions at the ward level. Model 1 includes indicators for (logged) median house prices at the ward level and at the local authority (LAD) level. To interpret these results, if we move from a ward where median house prices were £85,000 in 2014 to one where they were £230,000, support for Remain jumps by 10.5 percentage points. A similar shift at the local authority level results in an 8.1 per cent point increase in support for Remain. Models 2 and 3 repeat the exercise adding between 1996 and 2014 in ward median house prices. Moving from a ward with a ‘mere’ tripling of nominal house prices to one with a quadrupling is associated with a 5 per cent increase in support for Remain. The implication is that both long-run levels of house prices and relative growth over the past two decades shaped the Brexit vote. Places that had always done, or were recently doing, ‘better’ were happier with the status quo of remaining in the European Union. Figure 2 provides a striking example of this relationship, splitting out the voting pattern for wards in pro-Leave Wakefield, marginal Stockport, and pro-Remain Cambridge. In each district, there is a very strong relationship between wards with higher house prices having greater support for Remain, even though the three districts had very different average house prices (and indeed support for the EU). House prices both proxy for local economic fortunes and directly drive (dis)satisfaction with the status quo (since those in cheaper houses may resent their exclusion from the housing boom and those with highly appreciated houses have benefitted substantially). In both cases, these mixed fortunes appear to have filtered over into broader (dis-)satisfaction with contemporary British life. The obvious question to ask, given the relationship between housing and the Brexit vote, is whether it transferred over to voter preferences in 2017. The answer appears to be—only partially. That is, many richer Remain voting areas did back the nominally harder Brexit Conservative party in 2017—for example, in David Cameron's old constituency of Witney. Similarly, while the Conservatives had hoped to win many Leave voting Labour constituencies in the Midlands and the north, in reality only a handful of these seats switched sides. Figure 3 gives a sense of the degree to which the Brexit vote scrambled existing voting patterns and their relationship to the housing market. It plots, using the British Election Study of 2017, the probability of an individual voting for the Conservative party by two factors related to where they live—house prices in their local authority and support for Remain in their local authority (controlling for age, gender, education, income, and region). The three unbroken lines represent predicted probabilities of voting Conservative for individuals in areas where house prices average £750,000, £270,000 and £100,000, with the ‘whiskers’ on the plot showing the confidence interval around those predictions. The standard pattern where richer areas vote Conservative and poorer areas vote Labour broadly held up in 2017: voters in local authorities with more expensive housing tended to vote Conservative. However, it is notable that the ‘Remainness’ of areas was negatively connected to Conservative support and that was particularly the case in areas with expensive houses. The support for Conservatives in areas with cheap houses and high support for Leave failed to make up this gap. Graphically, we see a ‘narrowing’ in the effect of house prices on Conservative support as we move from pro-Leave to pro-Remain districts. Since more expensive areas were also more likely to vote Remain (and cheaper places for Leave) on aggregate the sharp drop-off in Conservative support in expensive pro-Remain areas, accompanied by a failure to pick up enough support in pro-Leave poorer areas, arguably lost the Conservatives their majority. In other words, the connection between housing and support for Remain, which cut across traditional class geographies, likely shaped the final outcome of the 2017 general election. Geographically then, house prices cut across traditional lines of political support. What about at the individual level? How did homeowners feel about the Conservatives and was there really a ‘rentquake’? The answer is mixed—it might be better, though awkwardly, described as a ‘rent plus mortgage-quake’. Returning to the British Election Study of 2017, we abstract away from geographical setting and look at individual level characteristics alone in explaining support for the Conservatives. In Figure 4, we examine the effect of housing tenure—conditional on age—on voting Conservative in 2017 (controlling for education, income, gender, Brexit vote, immigration attitudes, and employment status). The figure plots the predicted probability of an individual voting Conservative, with the grey line denoting renters and the black line denoting homeowners (with the confidence intervals around these predictions denoted by the dashed lines). Perhaps the most striking result to consider is that whereas among renters (in private and social housing) age is essentially unrelated to vote choice, among homeowners, there is a striking difference of around 30 per cent as we move from people in their twenties to people in their eighties. Put differently, young people behave very similarly regardless of housing status, whereas among the old, homeowners are much more pro-Conservative than private renters and social housing tenants. The relationship between age and partisanship among homeowners is a function of the variable conditions of their property ownership. For an older cohort, homes were bought cheap and held dear: they reaped the largest gains from the asset boom of the late 1990s and early 2000s, and have since transitioned into outright ownership. For a younger cohort, by contrast, homes were purchased at exorbitant price levels—and only by signing onto decades-long mortgages, which expose them to continuous risk of negative equity. Between 2005 and 2016, the percentage of mortgages issued that last thirty-five years or over rose from just 2 per cent to over 15 per cent. In other words, two mechanisms separate young and old homeowners: wealth and risk. Together, they constrained support for Conservatives among younger homeowners in the 2017 election—many of whom voted for a Labour manifesto with strong social protections—while driving it among older homeowners. As is well known, relatively few people own homes in their twenties at present. Still, the longstanding gap in preferences between owners and renters—around which Margaret Thatcher based her political strategy of Right to Buy—only emerges among the middle-aged. Homeowners under the age of around forty behave very similarly to renters politically. Since almost all of these younger homeowners will still have mortgage payments to make, we see here a new ‘coalition’ of renters and mortgagees, against those who fully own their property. Another way of viewing this is that those who benefitted most greatly from rising house prices, by virtue of having bought their house over a decade ago, were the core Conservative support base. One final regional dimension of the 2017 general election is worth noting. House price differences have dynamic effects on political behaviour in that they can ‘lock’ people out of expensive housing markets, forcing them to look for cheaper property elsewhere. There is some evidence that this pattern had political implications in 2017. London house prices rose much more dramatically in the post-2013 housing boom than did prices elsewhere. Accordingly, many younger workers and families moved into the wider south-east and east regions to seek cheaper property, often while continuing to work in London. As Ian Warren argued in a post-election analysis in The Guardian, many of these people carried their previous voting habits and social attitudes with them.11 Analysis of the British Election Study of 2017 shows that, controlling for age, education, gender, and income, individuals were most likely to switch towards Labour from 2015 to 2017 if they lived in the south-east and east regions, as opposed to London. Should this pattern continue, it will further scramble the longstanding political geography of Britain, again driven by the housing market. We have seen that housing is increasingly shaping the geography of British politics in novel ways. What about the possibility of reforms to the housing market itself? We see three key general areas for reform, each with striking political difficulties. The first is to lower prices by reducing the demand for housing relative to other goods. The second is to lower prices by increasing the supply of dwellings, both private construction and social housing. The third is a broader regional recalibration of the British economy that serves to reduce geographical inequalities in house prices. Reducing demand for housing essentially means reducing the growth of mortgage credit both for owners and landlords. This is probably the most advanced area of policy generation given the tighter restrictions on mortgage lending brought in by the Financial Conduct Authority in 2014, which required much more stringent accounting of borrowers’ income and expenditure and capped loan to income ratios. On the tax side, fiscal reforms have made buy-to-let substantially less attractive by increasing stamp duty and reducing tax deductibility. However, such policies lack an obvious political support coalition—landlords and would-be borrowers are directly negatively affected, owners in general may see lower prices, and the only obvious beneficiaries are future buyers—typically young and politically inactive. In terms of increasing housing supply, there may be some more obvious winners, but our analysis suggests that overall policy preferences are deeply divided along the tenure cleavage. Homeowners, eager to protect the value of their nest egg, tend to oppose new housing construction—especially when they expect house prices to rise in the coming years. Renters, by contrast, become increasingly supportive of social housing construction as their expectations of house price inflation rise. The divergent interests of these two groups, in other words, produce direct conflict over policy preferences. Some indicative results appear in Figure 5, which uses the British Social Attitudes Survey of 2010, asking people what their expectations are of future growth in the housing market and their attitudes towards the construction of social housing. The line with hollow circles represents the average predicted probability of supporting construction of social housing among homeowners, whereas the line with filled circles represents average support among renters (the whiskers represent the 95 per cent confidence intervals for these predictions). While attitudes of renters and homeowners are similar in declining or stable housing markets, in rising markets, the two diverge, with homeowners substantially less supportive, and renters more supportive, of social housing. Hence, building a cross-group coalition has been difficult during the secular rise of house prices over the past two decades. The support for social housing construction among renters reflects a significant change in ideal tenure. When house prices were low and stable, renters could expect to rise up the property ladder and move comfortably into homeownership. As house prices rise, however, renters are driven away from the desire to purchase an exorbitantly expensive home and toward a ‘social’ alternative that can protect them from the inflation of the private market. We find a strong and significantly positive relationship between renters’ sense of house price inflation and their desire to live in social housing. Su

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 57 citations exist upstream.