Considerable progress has been achieved in understanding public opinion on the environment since the turn of the 21st century (Prakash and Bernauer Reference Prakash and Bernauer2020: 1131). Much of this has been made possible due to the fielding of large-scale, cross-national surveys – such as the International Social Survey Programme (ISSP) and the World Values Study (WVS) – whose data have enabled researchers to examine environmental attitudes over time and across contexts (Birch Reference Birch2020; Dalton and Rohrschneider Reference Dalton and Rohrschneider2015; Franzen and Vogl Reference Franzen and Vogl2013; Kenny Reference Kenny2020). While such contributions have been extremely valuable, with much of the literature being cross-national in scope, it sometimes lacks close attention to the detail of the long-run development in national contexts and to changing sociopolitical conditions. US environmental attitudes have been extensively studied but, especially with it being an outlier in relation to its environmental policies (Harrison Reference Harrison, Harrison and McIntosh Sundstrom2010), calls have been made to look beyond this heavily studied case (Prakash and Bernauer Reference Prakash and Bernauer2020: 1131).
In this article I examine how environmental public opinion has evolved in Britain.Footnote 1 In many respects, Britain is representative of wider advanced industrialized democracies where environmental concerns have become part of the political discourse and most people think the government has a responsibility for environmental protection, yet there are still divisions on whether existing environmental laws are too strict or should go further (Dalton Reference Dalton2019). With the changing value prioritizations and issue conflict lines that such democracies – including Britain – have seen over recent decades (see Dalton, Reference Dalton2019: chapter 5), this article can further our understanding of the stability of/change in public attitudes and shed light on the over-time patterns that predict when environmental public opinion – and hence potential bottom-up pressure on policymakers – rises and falls.
The British case, moreover, has further specific reasons that make it particularly interesting. It has in some respects been pioneering. The 2008 Climate Change Act was one of the earliest pieces of legislation of its kind, with legislatures in several other countries replicating its innovative features (Averchenkova et al. Reference Averchenkova, Fankhauser and Finnegan2020), while in May 2019 Westminster became the first national parliament to declare an environmental and climate emergency (Rode Reference Rode2019). This only tells part of the story though, as Britain has had ‘a relatively high level of [climate] ambition but also policy instability and unevenness, succinctly characterized by Irene Lorenzoni et al. (Reference Lorenzoni, O'Riordan, Pidgeon, Compston and Bailey2008) as ‘hot air and cold feet’' (Lockwood Reference Lockwood2021: S41). Furthermore, as Neil Carter and Conor Little (Reference Carter and Little2021) demonstrate, Britain has had substantial variation not only in the saliency of environmental issues but also in the degree of consensus or disagreement in the party system at different periods. Having such system-level variation provides a fruitful context for examining changes in public opinion under different political circumstances.
Such a study is overdue, with one of the last dedicated examinations of long-run developments in British environmental public opinion having been carried out by Pippa Norris (Reference Norris1997) a quarter of a century ago. Data availability has proved an issue given that most trends do not go back very far. Many of the environmental topics asked about during the 1970s and 1980s had become dated by 1990 as they dealt with very specific, localized issues that were overshadowed upon the realization of the global connectedness of the issue (Dunlap Reference Dunlap2008: 7). This includes Britain, where work had concentrated on attitudes towards the countryside (Brook et al. Reference Brook1992), which are substantially different to other environmental attitudes (Norris Reference Norris1997: 320).
During the early decades, there was also a tendency to develop new measures, rather than use those that had already been constructed, validated and tested (Gifford and Sussman Reference Gifford, Sussman and SD2012: 66). This may partly have been due to an assumption that different environmental concern measures would produce comparable results – which proved to be untrue (Van Liere and Dunlap Reference Van Liere and Dunlap1981). Yet, as Axel Franzen and Sebastian Mader (Reference Franzen, Mader, Franzen and Mader2021: 63) remark, ‘Measuring environmental attitudes, comparing them between countries and observing them over time only makes sense if the measurement is reliable and valid.’ Furthermore, David Daniels et al. (Reference Daniels, Kraft and Kamieniecki2012: 462) argue that ‘each [environmental] measure has integrity and would be best examined on its own, rather than combined with other measures into indexes seeking to describe higher-order constructs’; hence combining various environmental measures over the decades is not ideal.
I overcome these limitations through focusing on a long-running question capturing whether respondents would prioritize environmental protection over the economy, made possible by the recent release of previously unavailable historical polling data from the 1980s. The analysis has two primary advantages. The first is methodological. It keeps the environmental topic the same, so one can track changes on the same type of environmental attitude. The construct validity of environmental–economic trade-off questions has recently been demonstrated in the British case (Kenny Reference Kenny2021a). Second, the question is substantially important. In a scenario where respondents cannot choose to further both economic growth and environmental protection concurrently, the question provides a good indicator of the strength of respondents' support (Birch Reference Birch2020; Kenny Reference Kenny2020). While certain environmental policies do provide economic co-benefits, attitudes on such a trade-off may be particularly policy relevant given the weak empirical evidence that ‘Green Growth’ can be fully achieved and the likelihood that policymakers may still have to curtail economic growth to avoid perilous global warming and other environmental damages (Gugushvili Reference Gugushvili2021).
At the aggregate level, I reveal that prioritization has regularly fluctuated, rising throughout the 1980s, falling dramatically in the early 1990s and recovering to its previous peak by 2001, before falling in the subsequent years and this trend being exacerbated following the financial crisis. Support recovered during the last decade, with a noticeably sharp increase between spring 2018 and winter 2019. I also find that environmental prioritization trends are not always aligned with environmental salience trends, with salience being used to denote its prominence in individuals' minds (Wlezien Reference Wlezien2005: 557), as operationalized by whether the environment is considered one of the most important issues facing the country.
At the individual level, over the course of the 1980s as the public was becoming more informed about environmental dangers, the positive association between environmental protection and educational attainment strengthened notably, with no evidence of political divides. However, while education remains important, left-wing individuals are now more likely than right-wing individuals to prioritize the environment, pointing to greater politicization.
I begin by providing a theoretical background and an overview of the development of environmental protection as an issue in Britain. I then outline the data, move on to analyse both aggregate-level trends and changes in individual-level predictors, and conclude by discussing the implications.
There are two particularly useful lessons from the comparative literature for when environmental attitudes can shift. First, attention tends to increase following focusing events such as unforeseen environmental disasters (Green-Pedersen Reference Green-Pedersen2019) and can be displaced when other issues come to the fore (Downs Reference Downs1972). This is substantially important, with evidence that positive shifts in publics' environmental attitudes are associated with increases in environmental (Bakaki et al. Reference Bakaki, Böhmelt and Ward2020) and renewable energy (Anderson et al. Reference Anderson, Böhmelt and Ward2017) policy outputs, while using both the change in salience and tone of climate change in the media as a proxy for public opinion points to a link between public demand and new climate policies (Schaffer et al. Reference Schaffer, Oehl and Bernauer2022).
Second, ‘broadly speaking, people are most agitated about the environment during periods of economic prosperity and least interested when economic recession draws attention back to materialist issues’ (Carter Reference Carter2001: 125). There is much evidence at the cross-national level that environmental attitudes are sensitive to changes in unemployment rates (Duijndam and van Beukering Reference Duijndam and van Beukering2021; Kenny Reference Kenny2020; Scruggs and Benegal Reference Scruggs and Benegal2012). Longitudinal studies from Germany and Sweden do, however, suggest that the association between economic cycles and environmental concern and saliency, respectively, is not as strong now as previously (Harring et al. Reference Harring, Jagers and Martinsson2011; Hartmann and Preisendörfer Reference Hartmann and Preisendörfer2021). Moreover, panel data analysis from the US and New Zealand finds that the decline in environmental concern following the 2008 financial crisis cannot be accounted for by changes in individual-level perceptions of the economy or respondents' own financial situation, and in the US specifically, subnational variation in economic conditions (Kenny Reference Kenny2018a; Mildenberger and Leiserowitz Reference Mildenberger and Leiserowitz2017). While these latter studies may suggest a spurious relationship between the two factors or an alternative mechanism other than increased economic insecurity through which recessions correlate with environmental attitudes – such as the environment being crowded out of national discourse by economic woes – a British question-order survey experiment does evidence a causal relationship between prioritizing climate change action and economic perceptions (Kenny Reference Kenny2018b). While I cannot test causal mechanisms in this article, the long timeframe enables me to observe descriptively whether changes in environmental prioritization in Britain are associated with changing economic conditions.
To provide the necessary context, before introducing the data, I now detail key events in the development of environmental protection as an issue in Britain.
Environmental protection was a low-salience issue at the beginning of the 1980s. No government administration in the 1970s or Margaret Thatcher's first two administrations had much interest in or understanding about it; it was rarely mentioned and it was low on the issue agenda (McCormick Reference McCormick1991). Any legislative change was mostly piecemeal and incremental (Lowe and Flynn Reference Lowe, Flynn and Mohan1989). The country had the reputation of the ‘Dirty Man of Europe’ for a combination of poor environmental standards and its obstructionist approach to European environmental directives (Rose Reference Rose1991). Together with Ireland, from 1976 to 1983 the UK consistently had the lowest numbers within the European Economic Community who thought it was very important to protect nature and fight pollution, with the figure always below 50% (Hofrichter and Karlheinz Reference Hofrichter and Karlheinz1990: 129).
The 1980s saw a shift in the salience of the environment for various reasons, including the discovery of global environmental issues, prominent food and water safety warnings, and economic development that increased the strain upon the local environment (Rootes Reference Rootes1991). High-profile disasters including the 1986 Chernobyl explosion and the 1989 Exxon Valdez oil spill acted as focusing events and spurred public concern (Carter Reference Carter2001: 125). Thatcher's Royal Society speech in September 1988 in which she warned that ‘we have unwittingly begun a massive experiment with the system of this planet itself’ provided a new top-down narrative and was credited with bringing environmental issues to the mainstream (Warrilow Reference Warrilow2017: 331). The ‘greenhouse effect’ and ‘global warming’ also gained prominence in the latter years of the decade, with UK news articles mentioning the terms increasing from 30 in 1987 to 275 in 1988 and reaching 1,538 in 1989 (Kenny Reference Kenny, Franzen and Mader2021b: 84). Environmental awareness is not a sufficient condition for prioritizing it, but it is necessary – which is why increasing knowledge from a low baseline could increase concern and why these events could be expected to have affected public opinion.
Subsequently, the surprise – and short-lived – surge of the Greens at the 1989 European Parliament elections provided a shock to the political system. Of the main parties, only the Liberal Democrats had an incentive to incorporate environmental issues into their platform to gain an electoral advantage (Carter Reference Carter2006). They pledged to include the environment as a central part of their policy platform (Rootes Reference Rootes1991: 43), and from their 1992 manifesto onwards would consistently devote substantially more attention to the environment than Labour or the Conservatives, thus allowing for limited politicization (Carter Reference Carter2006). Both Labour and the Conservatives did engage in surface-level greening with their respective 1990 White Papers ‘An Earthly Chance’ and ‘This Common Heritage’, but this was quickly tempered (Godfrey Reference Godfrey2012) and there was not a fundamental paradigm shift in environmental policy (Gray Reference Gray and Gray1995: 10). Subsequently, the early 1990s recession may have contributed towards a cooling off among the public. It is also noteworthy that the attention the press devoted to global environmental issues decreased sharply in 1991, with coverage in the aftermath of the 1992 Rio Summit being only a fraction of 1990 levels (Carvalho and Burgess Reference Carvalho and Burgess2005). Furthermore, there was an elite-driven counter-movement in the 1990s to undermine environmental science, with the UK second only to the US in the publication of environmentally sceptical books (Jacques et al. Reference Jacques, Dunlap and Freeman2008).
Environmental issues were a low domestic priority for Labour upon its return to government in 1997. While Tony Blair had an unprecedented agenda-setting role in international climate change diplomacy, ‘his enthusiasm [for climate change] did not progress to engagement with broader environmental issues and failed to result in sufficient policy action at the domestic level’ (Carter and Ockwell Reference Carter and Ockwell2007: 14). The modest domestic climate reforms that were introduced were centred on the British Climate Change Programme of 2000 that embraced an approach based on market-oriented policymaking instruments (Mildenberger Reference Mildenberger2020: 222–223). The ‘Big Ask’ campaign – launched by Friends of the Earth in May 2005 – would, however, play a key role in raising public awareness of climate change, with its demand for domestic action through a Climate Change Bill. Into this context of rising environmental salience, upon becoming Conservative Party leader in December 2005 David Cameron pursued a ‘Vote Blue, Go Green’ strategy by invoking environmentalism to modernize the party's image. This placed huge pressure on the Labour government to respond, leading to a period of rare competitive consensus on the issue, which most notably resulted in the cross-party 2008 Climate Change Act (Carter and Jacobs Reference Carter and Jacobs2014).
The coalition government of 2010 saw the environment become a positional issue, with the Conservatives actively deprioritizing it while their Liberal Democrat partners fought to keep it on the agenda (Carter and Clements Reference Carter and Clements2015). This corresponds with individual-level divides. While 2005 British Social Attitudes (BSA) data demonstrated that partisan divides on environment issues were narrow, a 2011 YouGov survey showed that the magnitude of such divides had increased notably (Clements Reference Clements2013). And the latter part of the decade saw the environment become more central, with the occurrence of severe floods, climate strikes and Extinction Rebellion protests.
While a brief overview, a key issue has been that – apart from Ipsos Mori's series capturing whether respondents list the environment as one of the most important issues facing the country (starting in late 1988) – the lack of long-run trend data has limited our ability to analyse systematically whether and to what extent the public's views on environmental protection altered during the past four decades and whether the predictors of support have remained stable or changed. In the next section, I detail the data sets that I utilize to carry out this analysis.
I focus on a binary dependent variable that captures a trade-off for respondents between prioritizing environmental protection over economic growth (1) or vice versa (0). At the aggregate level, environmental responses are presented as a percentage of all respondents who took the survey. In the individual-level analysis, respondents who did not provide one of the two answers are treated as missing and analysis is restricted to those aged 18 and over.Footnote 2 I outline below the data sets I use, with detailed bibliographic information located in Online Appendix A.
I draw upon three slight variations that are summarized in Table 1. For the first variation, individual-level data from October 1982 from the Eurobarometer and from May 1985 and October/November 1988 from Gallup are used. Meanwhile, aggregate-level reports – in the absence of having the individual-level data – are taken from the first quarter of 1992 from the Health of the Planet Survey and from April 2001 and April/May 2004 from Gallup blog posts.Footnote 3
Additional individual-level data are taken from December 2005 from the WVS and from February–July 2018 from the European Values Survey (EVS) with a slight alteration to the response categories to include a reference to jobs. Further individual-level data are taken from October–December 2019 from the Pew International Science Survey, where there is a slight wording variation by adding the prompt ‘even if neither is exactly right’. One would nevertheless expect the questions to capture the same sentiments.Footnote 4
Given the small variation in wording from 2004 to 2005, a Eurobarometer question focusing on a binary trade-off between environmental protection and economic competitiveness from November 2004 and November/December 2007 is used to approximate whether any change is an artefact of question wording.
Finally, to compare with a different trend, I draw upon responses from Ipsos Mori's long-running monthly question on the most important issues facing the country.
For the individual-level analyses, my independent variables are key socio-demographic and political variables that are common across the timeframe and have been argued as being important for environmental attitudes (Gifford and Nilsson Reference Gifford and Nilsson2014). These are age, gender, educational attainment, employment status and political party preference.
Age is measured in brackets of 18–24, 25–34, 35–44, 45–54, 55–64 and 65+, gender as male/female, educational attainment as age finished/expected to finish education (14 or under, 15, 16, 17–18, 19–20, 21+) and employment status as employed, unemployed, retired, home duties, student and other. The operationalization is influenced by ensuring they are standardized across surveys. For instance, highest educational attainment would be more precise than age completed, given changes in the relative meaning of the age at which education was completed over the 20th century, with the length of compulsory education increasing and barriers to access in Britain decreasing (Breen et al. Reference Breen2009), but this is not present in the 1980s surveys.Footnote 5 Political party preference is captured using general election vote intention, though for 2018 I have to use which party appeals to respondents and for 2019 the party respondents feel closest to.
Figure 1 displays the aggregate trend for choosing to prioritize environmental protection. There is a sharp increase in prioritization over the 1980s. While in 1982 just 51% would prioritize the environment, this rose to 58% in 1985 and to 70% in 1988. The increase between 1985 and 1988 is similar to the increase between 1987 and 1989 in a BSA question on whether the countryside should be protected from development even if this sometimes leads to fewer jobs, for which pro-countryside responses rose from 60 to 72% (Brook et al. Reference Brook1992). This rapid change in a short period coincides with environmentalism permeating into the political mainstream, with the 1988 survey fielded just one month after Thatcher's Royal Society speech. With individuals being exposed to messages of the damage being done to natural surroundings and the previously accepted inherent benefits of economic growth being challenged (Inglehart Reference Inglehart1981: 895), these results demonstrate an increasing willingness to put environmental protection above economic needs.
By the first quarter of 1992, this had substantially dropped, with just 56% being willing to prioritize the environment – the lowest support seen among 11 high-income countries surveyed (Dunlap et al. Reference Dunlap, Gallup and Gallup1993: 135). Economic conditions could be expected to have been important. In this regard, between the 1982 and 1985 surveys, the country's seasonally adjusted unemployment rate had remained static at between 11 and 12% (Office for National Statistics 2020), though had dropped to just over 8% by November 1988 in tandem with increases in environmental prioritization. With Britain entering into recession in 1991 for the first time in a decade, this trend had reversed and it correlates with the changing responses to the environmental question. Evidence of the economy crowding out environmental issues during this period can be seen from the 1992 general election campaign, where environmental concerns were rarely mentioned and it was instead dominated by ‘bread and butter’ issues, including the poor performance of the economy (Norris Reference Norris1997: 323).
By April 2001 environmental prioritization had returned to 1988 levels. This was a few months after the deadly floods of late 2000 coinciding with the wettest 12-month period on record in England and Wales since 1766 (Kelman Reference Kelman2001) and in the midst of the foot-and-mouth epidemic (Woolhouse et al. Reference Woolhouse2001). This also followed a period of continuous decline in unemployment since its previous peak in 1993, reaching lows that had not been seen in almost 25 years (Office for National Statistics 2020). One subsequently sees gradual declines tying in with a broader trend of the environment losing salience across Europe between 2002 and 2003 (Benedetta and Vincenzo Reference Benedetta and Vincenzo2020: 8), falling to 67% in May 2004. Further decline is noted in 2005 with the economic growth and unemployment trade-off question showing prioritization at only 59%. While one may wonder whether this is an artefact of adding ‘jobs’ to the trade-off question in 2005, the 2004–2007 questions referring to prioritizing the environment over the competitiveness of the economy also recorded a decline from 65 to 61%. Though the 2008 financial crisis was subsequently attributed responsibility for reduced support for environmental policies, this trend is important as it demonstrates that support had begun to tail off beforehand while the economy was still in a prosperous condition, echoing US trends (Nordhaus and Shellenberger Reference Nordhaus and Shellenberger2009).
While there is unfortunately a gap in the series until 2018, data from the Pew Research Center (2010: 120) in which British respondents are asked the extent to which they agree with the environmental prioritization statementFootnote 6 indicate a decline of 7 percentage points in complete agreement from 2002–2007, no change from 2007–2009 and a 5 percentage point decrease from 2009–2010. Thus, this supports the downward trend in the binary variable over the first half of the decade, while indicating that environmental prioritization remained relatively robust to the initial shocks of the financial crisis, before dropping in 2010. This is similar to trends in supporting urgent action to tackle climate change which had remained steady from 2007–2009, dropped in 2010, remained stable until 2012 and recovered by 2015 (Kenny and Fisher Reference Kenny and Fisher2015).
Picking up the binary trend in 2018, this data entry shows aggregate responses were similar to those recorded in 2005/2007. This was fielded two months after the Blue Planet II finale, a widely watched BBC TV documentary series on the planet's oceans presented by Sir David Attenborough that was accredited with boosting support for tackling environmental problems, especially the ‘war on plastic’. It is thus perhaps surprising that the stated prioritization is not higher, though perhaps less so given that claims of the documentary's influence tended to be based on anecdotal evidence (Dunn et al. Reference Dunn, Mills and Verissimo2020). Yet a dramatic increase would take place over the coming 18 months when attention in public discourse greatly increased in light of Extinction Rebellion protests and climate-strikes, and the movement from a passive to a competitive consensus on the issue at the political party level (Carter and Little Reference Carter and Little2021). Not only would the increased prioritization reach highs not recorded before at 77%, but it managed an increase of approximately 20 percentage points in the timeframe of 1.5 years, something that had taken 6 years to achieve in the 1980s.
Furthermore, it is worthwhile comparing this with the more regular series of those listing environmental issues as one of the most important issues facing the country (also contained in Figure 1). Between late 1988 and early 1992, both environmental prioritization and environmental salience display declines, while between spring 2018 and winter 2019 both record increases. However, the only other period that has sufficient data-points of both series suggests that they may not always move in tandem. Between 2001 and 2007 prioritization is trending downwards, whereas salience is rather stable between 2001 and 2004, before then generally following an upward trajectory between 2004 and 2007 (with its peak in this period in the aftermath of the Stern review on the economics of climate change). One possible explanation is that measures of environmental salience – which can sometimes be regarded as ‘costless’ – may fail to convey the public's priorities (Norris Reference Norris1997: 326/327). However, without having the data-points to carry out a fine-grained comparison between the two series, one cannot make firm conclusions.
Changes in individual-level predictors
I now investigate whether the correlates of support for environmental prioritization have altered over time.
Table 2 displays the results of logistic regressions. The regressions in the 1980s allow us to see how predictors altered as environmentalism rose up the public agenda, whereas the 2018 data allow us to compare these 30 years on. At this stage, the 2005 WVS data is omitted because it is missing the political party preference question and the 2019 Pew data is excluded for not asking employment status or the age one completed their education.
Notes: * p < 0.05, ** p < 0.01, *** p < 0.001; coefficients are log-odds ratios; standard errors in parentheses. Lib = Liberal Party; LD = Liberal Democrats; SDP = Social Democratic Party; SNP = Scottish National Party; DK = don't know.
In 1982, there are three demographics with significant associations. Those over 65 are the least likely to prioritize the environment, while the greatest support is seen among the 35–44 age group. Those who finished their education aged 21+ display the greatest levels of prioritization relative to other categories, though there is no linear pattern in the lower groups. Lastly, women are less likely to prioritize the environment; while women are now usually more likely to display environmental tendencies, work in the 1980s found that gender differences were inconsistent (Gifford and Nilsson Reference Gifford and Nilsson2014). Partisan differences are notable by their absence; party support did not predict one's prioritization. This is in line with the limited politicization on the issue during the 1980s having not been heavily incorporated into the main political parties' platforms (Carter Reference Carter2006: 749).
By 1985, the almost 10 percentage-point aggregate-level increase comes with shifts in the demographic correlates. First, the substantial age differences disappear. The direction of the gender coefficient has shifted and is statistically insignificant. Education effects are only seen for those who finished at the age of 20 rather than for the 21+ group, though again there is a lack of a linear relationship. And there are no differences between supporters of different political parties. By 1988 – a point at which over 70% would prioritize the environment – the only significant correlate is educational attainment. Moreover, unlike previous years, each higher category of age completed education sees a higher magnitude of association. So throughout the 1980s, educational attainment is the only demographic correlate to be consistently related with prioritization, and its graded association has taken form in tandem with increasing attention to the issue.
How does this compare with correlates three decades later? There are similarities and differences. Educational differences remain, with the coefficients for finishing one's education at a later age being larger. There are again no gender differences, while those who are retired are less likely to prioritize the environment.Footnote 7 The major difference is in the relationship with party support. Though the 1980s surveys use vote intention while the 2018 survey uses party that most appeals to respondents, in 2018 those with a Liberal Democrat preference are much more likely to prioritize the environment than Conservative or Labour supporters, while Green Party supporters are even more likely still.Footnote 8
Another possibility for the lack of political divides in the 1980s could be that ideological divides existed at the bottom-up level but that these were not seen in voting patterns due to political parties not acting as vehicles for these, similar to what has been observed with class voting (Evans and Langsæther Reference Evans and Langsæther2021). One way to test this is by examining the role of left–right orientations and libertarian–authoritarian values, with the former largely concerned with equality and the latter with personal freedom (Evans et al. Reference Evans, Heath and Lalljee1996). While libertarian–authoritarian values are not present in the early surveys, the 1982, 2005 and 2018 surveys all have – in addition to left–right self-placement on a 1–10 scale – the four-item post-materialist index that correlates with these (De Graaf and Evans Reference De Graaf and Evans1996) and so can be used as a proxy. Materialists choose as their first two priorities maintaining order in the nation and fighting rising prices, postmaterialists giving the people more say in important political decisions and protecting freedom of speech, and those who choose one from each are mixed (Inglehart Reference Inglehart1971).
These regressions are displayed in Table 3, controlling for the same variables as in Table 2 with the exception of party preference, which was not asked in 2005. In 1982, there is no substantial relationship between left–right self-placement and environmental prioritization. However, postmaterialists are more likely to prioritize environmental protection than materialists. Thus, there was a group of individuals whose preferences were not being represented in the party system. In 2005, neither the postmaterialist nor the left–right coefficients were significant.
Notes: * p < 0.05, ** p < 0.01, *** p < 0.001; coefficients are log-odds ratios; standard errors in parentheses; controls for age, gender, educational attainment and employment status (see Online Appendix C for full table).
In 2018, however, left–right orientations were significantly and substantially associated with environmental prioritization, evidencing increasing politicization. Keeping everything else at their mean value, Figure 2 displays predicted probabilities for prioritizing environmental protection at different responses on the left–right scale. Furthest to the left one has a 75% probability of choosing the environment whereas for those furthest to the right it is just 46%. That this relationship was not present in 2005 or indeed in the 1980s suggests this is a recent phenomenon. In Online Appendix C, party preference is added to the models for 1982 and 2018 and the findings for left–right self-placement and postmaterialism remain robust.
In this article, I examined the long-term evolution of Britons' environmental prioritization. The aggregate-level trends show a notable degree of volatility. While rising during the 1980s, prioritization declined rapidly in the early 1990s, recovering by 2001 and entering into a gradual decline over the coming decade. Though levels of prioritization were the same in early 2018 as before the onset of the financial crisis – between which we know that they dipped further – by the end of 2019 they had risen to levels not previously recorded. Putting this together with data on environmental saliency, the evidence supports the view that public environmental sentiment at the end of 2019 has been at its highest since the end of the 1980s.
At the individual level, two results stand out. One is the consistency of educational attainment, with prioritization greatest among those who completed their education at a younger age. While age completed education was used for comparative reasons, university education has been demonstrated elsewhere to be particularly important (Cotton and Alcock Reference Cotton and Alcock2013). The second result of note is the increasing politicization. In the early 1980s and the mid-2000s, there was no evidence of a significant correlation between left–right self-placement and environmental preferences controlling for other factors, while by 2018 left–right self-placement had become important. Moreover, while from 1982 to 1988 individuals' political party preference was not related to their environmental preference, in 2018 and 2019 both Green Party and Liberal Democrat supporters are more likely to prioritize the environment than supporters of other parties. Thus, the evidence suggests that the politicization of the environment that has been documented at the elite level in Britain over the past decade is also present at the individual level (Carter and Clements Reference Carter and Clements2015).
What lessons can we take from this? First, we can see patterns in when environmental prioritization rises and falls. The high points of environmental prioritization occurred in the aftermath of significant environmental events such as Thatcher's 1988 Royal Society speech, the floods of 2000 and the 2018/2019 climate change protests. The recessions of the early 1990s and 2008 were associated with declines, though in the latter case it seems to have exacerbated a previous trend. Moreover, environmental prioritization and salience do not necessarily go hand in hand, evidenced by the increased salience of the environment from 2004 to 2007 while prioritization was subsiding. This parallels with a comparative study of public opinion in the US, Italy and China, where attention to climate change decreased during the COVID-19 pandemic, though public worry did not (Sisco et al. Reference Sisco2020). Thus, it cannot be presumed that trends in different facets of environmental public opinion will move in the same directions.
What is key is that the British public's prioritization of the issue has proved relatively fragile and has shifted dramatically. When prioritization is high, environmental activists and policymakers should not miss the window of opportunity to push for legislative changes that may at other times be more difficult to achieve. It is possible that recent levels of prioritization may prove more enduring, but this should not be taken for granted. Indeed, analysis of this trend in the US shows that between 2020 and 2021 – in line with increasing unemployment – Americans' willingness to prioritize the environment decreased for the first time since the 2008 recession (Saad Reference Saad2021). Such evidence points to support still being susceptible to downward shifts in response to economic conditions. Moreover, should the observed issue politicization in Britain endure, this may make the achievement of the consensus that was a hallmark of the 2008 Climate Change Act more difficult. Future research should pay careful attention to examining whether the environmental values and preferences evident in 2019 prove to be durable over the coming years or fade away.
The supplementary material for this article can be found at https://doi.org/10.1017/gov.2022.17.
I would like to thank Ben Clements, Peter Enns, Will Jennings, Joan Madia, Andra Roescu, Riley Taiji, Giacomo Vagni and the three anonymous reviewers for their constructive comments on previous drafts that greatly improved the article. An earlier version of the article was presented at the 2020 Environmental Social Sciences workshop at the University of Bern, and I am grateful for feedback received at this. This work was supported by the Economic and Social Research Council (grant number ESRC ES/S000380/1).
Permission to reproduce material from other sources
Pew data used in analyses: Pew Research Center bears no responsibility for the analyses or interpretations of the data presented here. The opinions expressed herein, including any implications for policy, are those of the author and not of Pew Research Center.