Sentience and salience – exploring the party politicization of animal welfare in multi-level electoral systems: Analysis of manifesto discourse in UK meso elections 1998–2017

ABSTRACT This study explores the party politicization of animal welfare in the context of multi-level governance in the UK. It examines over 1300 pledges in party manifestos for Scottish, Welsh and Northern Irish elections 1998–2017. It reveals the nature of party competition, including increasing salience over electoral cycles. This is complemented by examination of the party dynamics in two sub-fields: wildlife protection and farm animal welfare. The wider significance of this study lies in showing how the move to multi-level electoral politics provides new political spaces to advance animal welfare and how meso-ballots are increasingly attuned to the symbiosis of humans and animals. These factors are driving the territorialization of policy and leading to distinctive animal welfare regimes in the different countries of the UK. In addition, partisan theory reveals how the electoral politics of animal welfare varies between sub-fields and is shaped by parties’ relationships with different policy communities.


Introduction
Over recent decades animal welfare has become a prominent issue in many Western democracies (Hoy-Gerlach et al. 2019;Lundmark et al. 2014;Sunstein and Nussbaum 2006;Vogeler 2017). As an umbrella term it encompasses a broad, sometimes contested, range of norms and policy measures designed to ensure the well-being of non-human species. Drawing on partisan theory, that underlines how political parties may act as vectors advancing differing policy preferences into policymaking (Hibbs 1977;Vogeler 2019), this study Accordingly, the principal aim of the following analysis is to explore how the issue of animal welfare has been addressed in the new political space provided by 'devolved' or meso-elections in the UK. As an exploratory study, we first take a holistic view of the electoral politics of animal welfare before engaging with partisan theory and exploring the contrasting party dynamics of two sub-fieldsas set out in the following research questions: (1) At an aggregate level what are the principal characteristics of animal welfare pledges in UK meso elections in terms of topics, party politicization and issue-salience? (2) What are the principal characteristics of meso-electoral competition in relation to the sub-fields of: (a) farm animal welfare and (b) wildlife protection?
Our main argument is that, hitherto largely overlooked by scholars, the move to multi-level electoral politics provides new political spaces to advance animal welfare. Our key findings are that: meso-ballots are increasingly attuned to the symbiosis of humans and animals and that this is driving the territorialization of policy and leading to distinctive animal welfare regimes in the different countries of the UK; and the electoral politics of animal welfare varies between sub-fields and, as suggested by partisan theory, this is shaped by parties' relationships with different policy communities. Accordingly, the remainder of the article is structured as follows: first, the research context and study methodology are outlined. In the following sections the findings are presented in relation to each of the three research questions. The article concludes with a discussion of the principal findings and their implications.
Research context and social theory (i) The salience of animal welfare Humans and animals are inextricably tied in a complex set of relationships (Donaldson and Kymlicka 2011). These are not static phenomena, for as Macnaghten (2001, 4) states, over the past century they have 'undergone a profound set of transformations … [Hitherto] animals figured in the modern project principally as resource for human progress … [Latterly] the subordination of animals is being seriously questioned'. This shift is reflected in public attitudes data that indicate support for policy measures to promote animal welfare (Deemer et al. 2010). As the following indicates, 'animal welfare' is an umbrella term that is made up of a series of sub-fields associated with: farm animal cruelty cases (Weary 2018); the environmental movement and calls to tackle habitat loss and climate change (Beebee 2018;Butterworth 2017); animal testing of drugs, cosmetics and other products (Monamy 2017); debates around human consumption of meat (Rachels 2012); hunting and animal involvement in 'sports' (Woods 2003); intensive livestock and poultry farming techniques (Cornish, Raubenheimer, and McGreevy 2016); and the use of fur and animal products in garments and other artefacts (Makarem and Jae 2016). These diverse sub-fields have all seen political activism over animal rights and welfare (Cherry 2016). In turn, the level of activism has been shaped by increased media attention to the issue (Rodan and Mummery 2016); as well as a concern to uphold legal protections, including the recognition of animals as sentient beings in the Treaty of Amsterdam, (European Communities 1997).
(ii) Theorizing animal welfare, electoral politics and policy change A full discussion of all the different and contested notions of 'animal welfare' is beyond the present purposes (see for example, Carenzi and Verga 2009;Mellor 2016). As Carenzi and Verga (2009, 22) observe, it is a 'multi-faceted issue which implies important scientific, ethical, economic and political dimensions'. For the present purposes it refers to the avoidance of the negative feelings and experiences related to the 'Five Freedoms' (freedom from: thirst, hunger and malnutrition; discomfort and exposure; pain, injury, and disease; fear and distress; and freedom to express normal behavioursee McCulloch 2013; Webster 1994)as well as measures to uphold more sophisticated conceptualizations concerned with sentience, the generation of positive and subjective animal experiences and, improving human-animal relations (see Mellor 2016).
Extant studies of animal welfare fall into three categories, namely: 'pan animal welfare' studies that treat animal welfare as an holistic issue (Jukan, Masip-Bruin, and Amla 2017;Mellor 2016); those that offer a discrete focus on subfields (such as animal experimentation, hunting, transportation and so on); and hybrid work that combines the foregoing categories. This is a hybrid study. As the initial exploratory study of animal welfare in UK mesoelections it is appropriate to first take an holistic, pan-animal welfare view (inter alia, this gives us new insights into how the salience and nature of animal welfare pledges have developed over time and between polities and how it compares to other policy issues in party programmes). We then examine the electoral politics of two sub-fieldsfarm animal welfare and wildlife policy. Informed by partisan theory, this allows more sophisticated analysis of the nature and extent of contrasts between sub-fields.
The literature on public policy offers a raft of theories to understand policy change (Van der Heijden and Kuhlmann 2018) that can be applied to the present examination of the electoral politics of animal welfare. A core debate centres how policy change links to the politicizing of policy issues (Beland and Waddan 2012). Politicization is reflected by rising public salience, contestation and, crucially, engagement with a growing number of policy actors (Börzel and Risse 2009). It results in issues becoming the subject of political competition with different policy actors mobilizing to advance their preferences and secure policy change by pressurizing government to address their demands. In this regard, partisan theory underlines the key role that political parties play in acting as vectors that advance differing policy preferences into policymaking. According to this view they may pursue different policies for a range of motivesincluding policy-seeking to promote the policy preferences of party members, or vote-seeking for electoral gain (Hibbs 1977;Vogeler 2019). As partisan theory reveals, parties' electoral discoursethe nature of their manifesto pledges, is shaped by their engagement with a growing number of policy actors. As noted, the following analysis selects two-sub-fields where the extant literature suggests there will be contrasting party dynamics.

Methodology
We operationalize our analysis of partisan theory and animal welfare policy by examining the party politicization of the issue in the party manifestos in UK meso-elections. Party politicization refers to the process by which a policy issue develops to be electorally salient, thereby rising up the political agenda and becoming the subject of inter-party competition (Carter 2006). 'Issue-salience' is the indicator used to assess competition in manifesto studies. It refers to the importance of an issue for a given party in a particular election as indicated by the number of policy proposals (RePass 1971). Thus analysis of party election programmes can reveal inter-party politicking as each attempts to advance their policy demands and secure voters' support. Following established practice (Topf et al. 1994), in the following analysis issue-salience is determined by content analysis of manifestos (Neuendorf 2002). Accordingly, electronic versions of the manifestos of the leading 1 parties in all 'devolved' elections from 1998 to present were analysed using appropriate software. 2 Following established practice in electoral studies, the manifesto texts were divided into 'quasi-sentences' (or, 'an argument which is the verbal expression of one political idea or issue', Volkens 2004, 96). Splitting sentences in this way controlled for long sentences that contained multiple policy proposals. 3 Subsequently, the quasi-sentences were tagged using a coding frame based on key topics or sub-fields derived from the policy literature on animal welfare (e.g. animal experiments/ vivisection, farm animal welfare, hunting, trade in rare and protected species, and so on). In addition, all the quasi-sentences were coded using Beth Reingold's notion of 'direction'. Originally used in the study of parliamentary debates and whether interventions were deemed to be feminist, anti-feminist or neutral (Reingold 2000, 166-177); here it is applied to determine whether manifesto pledges are proor antianimal welfare. The overwhelming majority were found to be pro, with a handful (<1.0 per cent) that were 'anti' (such as pledges advocating the repeal of hunting bans). 4 Our research design is based on a two-stage approach. 1. As an exploratory study looking at the issue of animal welfare in UK meso elections, we first undertake holistic, cross-polity analysis of parties' records on animal welfare. 2. Then, following leading research (Lundmark et al. 2014(Lundmark et al. , 2018Vogeler 2019) highlighting the way political parties' propensity to advance animal welfare pledges is shaped by their responsiveness to different exogenous interests (lobby groups, campaigners, business interests, unions and media companies), we select two key sub-fields (farm animal welfare and protecting wildlife) and examine party politicization in relation to each. This allows us to understand the party politicization of the sub-fields within the wider disciplinary context of animal welfare as whole.
In methodological terms, a further consideration that had to be addressed was the best way to measure the changing issue-salience of animal welfare over time. Here it should also be noted that, as a broad generalization, over recent decades manifesto word lengths have increased as party programmes have become more comprehensive (Dale 2000;Däubler 2012). Yet, as the corpus of manifesto studies confirms, issues such as animal welfare do not inevitably gain more pledges over time as a result of the expansion of the topics covered in party programmes. Instead, the number of quasi-sentences on different issues is volatile, rising and falling over election cycles. Furthermore, whilst the long-term trend may be one of increasing manifesto length, empirical analysis shows word lengths to be highly variable (decreasing as well as increasing) over electoral cycles. The meso-elections studied here are no exception. 5 For this reason, this study follows the approach used by a broad raft of electoral studies and analyses the absolute number of animal welfare manifesto pledges. 6 However, for additional rigour comparative measures of issue-salience are also provided showing parties' attention to animal welfare in relation to other policy topics and in comparison to state-wide manifestos.
The following analysis also explores the Left-Right party dynamics animal welfare pledges. In methodological terms this presented the challenge of determining the Right-Left orientation of each party. In response, this was decided by following each party's self-described position, as set out in party literature and websites. Thus, the Conservative and Unionist Party, Ulster Unionist Party, Democratic Unionist Party, UK Independence Party are designated Right-of-centre. The Alliance Party of Northern Ireland asserts that it is a cross-community centrist party. The remainder are Left-ofcentre. However, as a dedicated literature explains, classifying civic nationalist parties according to a Left/ Right dyad can be a contested matter. Whilst most accounts determine that the Scottish Nationalist Party and Plaid Cymru are Left-wing (Massetti 2018), others point to a more chameleon-like political identity; whereby both parties are prepared to advance pledges designed to appeal to conservative, often rural constituencies, whilst at the same time espousing Leftist/ centrist pledges on welfare and public services (Massetti and Schakel 2015; Wyn Jones 1996). In the following analysis we followed the SNP's and Plaid Cymru's self-stated orientation. For its part the former describes itself as 'Scotland's largest political party and party of Government. Centre left and social democratic'. 7 Whilst Plaid Cymru (2016, 4) describes its core aims as 'to secure independence for Wales in Europe [and …] to ensure economic prosperity, social justice and the health of the natural environment, based on decentralist socialism'.

Research findings
This section is comprised of three parts. As noted, an initial focus on parties' attention to animal welfare in their meso-election manifestos is followed by an examination of partisan differences in relation to two sub-fields: wildlife protection and, animal welfare.
(1) At an aggregate level, what are the principal characteristics of animal welfare pledges in UK meso elections in terms of topics, party politicization and issue-salience?
Using both absolute and relative measures, the present analysis shows how, as devolution has matured, animal welfare policy has been subject to increasing issue-salience in meso-level electoral competition ( Figure 1). In absolute terms, over successive electoral cycles there is a significant increase in the salience of animal welfare across polities. Under a fifth (19 per cent) of the total pledges 1997-2019 were made in the first two election cycles in each polity compared to over a half (54.6 per cent) in the last two election cycles. The total number of meso-election pledges rises from 146 in the 1998/1999 ballots to 646 in the 2016 elections. 8 The greatest rise in salience is in Welsh elections. This reflects the more limited animal welfare related powers in the pre-2006 devolution settlement for Wales (Figure 1(c)). Scotland and Northern Ireland have similar rates of increase over electoral cycles (Figure 1(b,d)). The data also reveal how the salience of animal welfare varies over cycles; declining in specific ballotsmost noticeably in 2011. In part, this temporary decline may be explained by policy gains made over the preceding five years. 9 The Scottish data also demonstrate the disproportionate influence of the Green Parties in the meso-ballots (see below) for the sharp decline in the salience of animal welfare in the 2011 Scottish election is in large measure due to the limited attention that the party gives to the subject compared to previous and subsequent ballots. In conceptual terms these fluctuations in salience underline the applicability to the meso-electoral politics of animal welfare of what policy scholars term Punctuated-Equilibrium Theory (True, Jones, and Baumgartner 2007). This means that fluctuating levels of issue-salience are contingent upon (and reflective of) emerging debates and specific issues. In the period studied, these include: an increase in the salience of farm animal disease policy in the 2003 elections (in response to the 2001 Foot and Mouth disease outbreak in the UK) and subsequently, media attention and NGO activism in relation to animal welfare and live exports, the use of animals in circuses and poor regulation of animal testing, as well as increased devolution of animal welfare powers. Relative indexes also confirm the increasing salience of animal welfare in the meso-elections. When animal welfare pledges are expressed as a percentage of the number of pledges made in two of the main devolved policy areas: health and education (Figure 2), animal welfare rises from under a quarter to a third of the comparison policy areas. In this regard the mesotrend can be seen to mirror the increasing salience of animal welfare in state-wide elections (Figure 1(d)). This matters because it provides empirical evidence of how in countries with rising public concern over animal welfare such as the UK (Ipsos/ Mori 2015), there is a marked increase in the issue-salience of animal welfare in party manifestos in both state-wide and meso-elections. In turn, this confirms that the new political spaces provided by devolved elections are providing new opportunities for parties to seek voter support for territorial policies on animal welfare.
At this point it is helpful to explore parties' stated motives for advancing animal welfare measures in their election programmes. A core trope is the connection between the mistreatment of animals and violent crime. For example, 'We recognise the links between animal cruelty and human violence and believe there should be formal recognition of these links' (Scottish Green Party 2007a, 3) and, 'Scottish Labour recognises that cruelty to animals is an indicator of the wider prevalence of violence in society' (Scottish Labour 2007, 7). Others underline the link between the economy and well-being, 'wildlife and habitats -Biodiversity has an important role to play in the social and economic life in Scotland, contributing to the wellbeing of Scotland through access to wildlife-rich environment' (Scottish Liberal Democrats 2007, 11). Generativity is a further motivating factor, Given our capacity to destroy, we have a responsibility to protect and conserve those other species with which we share the planet, and to leave a planet for future generations that is no less diverse than the one that we inhabit today. In addition, an important, if less prominent reason is the impact of individual dietary choices, 'Equally, the future of the environment and the rights of animals should not be prejudiced by the way we feed ourselves' (Scottish Green Party 2003, 7).
As noted, animal welfare covers a broad range of sub-fields. Overall, across devolved polities wildlife protection/ biodiversity was the lead issue or subfield in the electoral discourse (34.3 per cent of quasi-sentences). This reflects the policy traditions of the earlier era of administrative devolution when such matters were devolved to the old territorial ministries of the British government: the Scottish, Welsh and Northern Ireland Offices. The second most salient issue in the post-1998 manifesto discourse is regulation/ criminal justice measures (26.3 per cent). This is significant. It shows discontinuity and parties' intention to use the new policy levers available to the devolved governments to move beyond the limited autonomy of administrative devolution and establish new territorial animal welfare regimes. The third-placed sub-field was farm animal welfare (8.9 per cent).
The present analysis also reveals how party programmes on animal welfare have become more comprehensive (spanning more issuesor sub-fields) over successive election cycles. Across polities the percentage of party pledges outside the core three areas of farm animal welfare, protecting wildlife/ biodiversity, and regulation, increased from 25 per cent in the 1998/1999 elections to 35 per cent in 2016/2017. Again, this is significant as, over time, it reflects an increasingly sophisticated understanding of the factors that constitute animal welfare by parties and issue public. It is also indicative of greater devolved legislative and policy competency on animal welfare as more powers are transferred from Westminster. It is evidenced by the fact that pets, hunting and testing/experimentation receive no attention in the 1998/1999 elections. Yet along with 'sport', slaughter, promoting animal welfare overseas and circuses, these issues become established areas of inter-party competition in subsequent elections.
A further key finding is how devolution in the UK is characterized by divergence in the issue-salience of different animal welfare topics between polities. In terms of the total number of quasi-sentences, animal welfare is a significantly more salient election issue in Scotland and Northern Ireland compared to Wales. The majority of pledges were made in Scottish elections (43.7 per cent), followed by Northern Ireland (35.5 per cent) and Wales (20.9 per cent) ( Table 1). Differences in devolved competency on animal welfare; shorter, less detailed Green Party manifestos and relatively low salience in Welsh Labour Party programmes (compared to Scottish Labour) explain the difference between Wales and the other polities.
Existing studies underline how asymmetrical devolution -(in other words, the situation when meso-parliaments in quasi-federal states do not have uniform powers)may drive the territorialization of policy and law (Brown  2011,17). The Scottish discourse also places greater accent on 'sport' and regulating pets. For example, 'we will toughen up regulations with regard to the welfare of racing greyhounds' (Scottish Labour Party 2011, 9) and, we will 'require the current working group on pet welfare to bring its work to an early conclusion to allow for updating of the laws on pet welfare in order A further key finding is that quasi-federal governance in GB is leading to variation in the issue-salience of animal welfare between territorial divisions of individual state-wide parties ('intra-party variance') (Table 3). Scottish Labour makes over three-quarters of the Party's meso-election animal welfare pledges (1999-2016) (76.5 per cent); Welsh Labour under a quarter. Thus, for example, Scottish Labour gives more attention to hunting/ culling, sport and pets. Reflecting the asymmetrical nature of devolution in the UK (criminal justice has yet to be devolved to Wales, but is devolved to Scotland), the greatest divergence between the Scottish (25.2 per cent of pledges) and Welsh (0.9 per cent) divisions of the Labour Party relates to 'Regulation/ Criminal justice'. For example, 'Scottish Labour believes that more action is needed on wildlife crime and we will ensure that the move to a single police force is used as an opportunity to deliver concerted action across Scotland to tackle this issue' (Scottish Labour Party 2011, 11).
For the Conservatives, the Welsh wing of the Party gives greater attention to farm animal welfare than its Scottish counterpart (10.3 compared to 4.1 per cent of all Conservative pledges). The same is true of policy on animal slaughter (6 per cent compared to zero). In contrast, the Scottish Party predominates on hunting and pets. For example, 'As it is so widely resented in rural communities as an unwarranted attack on their way of life, we will also allow Parliamentary time for a debate on the repeal of the Wild Mammals Act' (Scottish Conservative Party 2003, 7).
In the case of the Liberal Democrats, the Welsh Party predominates on farm welfare and trade. For example, 'Welsh Liberal Democrats will … Develop the use of individual farm health plans' (Welsh Liberal Democrats 2003, 9). Whilst the Scottish Party places greater emphasis on pets. For example we will 'Introduce a Protection of Animals Bill to allow local authorities to take into their care animals at risk of abuse' (Scottish Liberal Democrats 2007, 8).
The significance of these findings is in underlining the discontinuity introduced in 1998/1999 when the UK moved away from sole reliance on statewide (Westminster) elections and introduced three meso-electoral systems. This has major implications for the state-wide parties. It presents opportunities for policy innovation, yet it also raises challenges in terms of consistency of policy positions, notions of territorial justice and the public's perception of each party. Crucially, it introduces a new level of accountability  (2) What are the principal characteristics of meso-electoral competition in relation to farm animal welfare?
The present analysis of meso election pledges reveals a cross party desire to tap rising public support for farm animal welfare at the ballot box (Clark, Stewart, and Panzone 2016). The Green Party made most pledges on the issue (39 per cent of all party total of pledges, N = 130; or 10 per cent of the party's animal welfare pledges), followed by the Liberal Democrats (13 per cent of all party total of pledges) (e.g. We will: Develop a national strategy, co-ordinating the agricultural industry, the veterinary and emergency services, to respond to, and tackle animal diseases ', Welsh Liberal Democrats 2007, 6). In turn, the they are followed by the nationalist parties (12 per cent), Conservatives (11 per cent; e.g. 'Welsh Conservatives will lead the way in ensuring better protection for the nation's animals, whether they be companion, farm or wild', Welsh Conservatives 2016, 17), and Labour (10 per cent; e.g. 'Scottish Labour supports the rights of landowners and farmers to sensibly manage their land, the wildlife on it and livestock. We remain firmly on the side of the majority of farmers and landowners who believe this should be done with … animal welfare as a core concern', Scottish Labour Party 2007, 14). Green Party predominance on this issue is no surprise and reflects a sophisticated understanding of the importance of sustainable ecosystems. As Eckersley (1992, 34) notes the Greens' principal concerns are 'resource conservation, human welfare ecology, preservationism, animal liberation and ecocenticism' (see also Dobson 2012). In the meso elections Green Party pledges give particular attention to proscribing intensive confinement of farm animals. For example: 'We will immediately abolish all battery and enriched egg production systems and replace them with free-range systems' (Scottish Green Party 2003, 8); and we will 'Phase out the sow stall system' (Green Party of Northern Ireland 2003, 7). The absence of a significant Left-Right cleavage between the two state-wide parties on farm animal welfare is another key finding (14.9 per cent of Labour's animal welfare pledges in the meso-elections were concerned with farm animal welfare, compared to 14.4 per cent of Conservative Party pledges).
Reference to the extant literature allows us to place these findings into context with reference to partisan theory. Thus, Tosun (2017Tosun ( , 1623 reveals how 'green parties in Germany and the United Kingdom are the strongest advocates of opening up policy-making in the agri-food sector to new actors and of proposing policies that span across the boundaries of this policy sector'. Similarly, Vogeler's analysis (2017) reveals partisan differences on farm animal welfare across several European countries. This work tells us that generally, conservative parties often have close ties with farmers' unions and may therefore oppose farm animal welfare regulations. This is also supported by research in the USA (Hurwitz, Moiles, and Rohde 2001, 911) that noted how Right-of-centre (Republican) party dominance in Congress shaped voting regulation and was in turn, 'strongly related to the degree of agricultural interests in a member's district'.
Whilst historically there have been strong links between the Conservatives and farming unions, over recent decades the UK emerges as an exception to the wider international trend. For example, Chaney (2014a) found that when the level of attention to all animal welfare sub-fields was compared in parties' manifestos for Westminster elections, the top two issues with greatest crossparty consensus were farm animal welfare and vivisection. In terms of issuesalience, farm welfare pledges were ranked third in the Conservative Party's top animal welfare issues. Analysis of the 2019 UK state-wide elections by campaigning animal welfare interest groups also points to Conservative Party receptiveness to farm animal welfare interest groups' policy claims on strengthening farm welfare policy and law. As one campaigning group revealed, there has been particular progress on pledges prohibiting live exports, banning low-welfare imports, introducing further subsidies for farm welfare, and legal recognition of animal sentience (CIWF 2019, 1).
The Right-of-centre Conservative Party's championing of farm welfare policies can be explained in the following terms. Whilst partisan theories contend that parties can exert an independent influence on voter behaviour and policy outcomes (see Aldrich 1995), Rohde's (1991) conditional party government theory argues that when there is a changing environment in the electorate, notably shifting public attitudes on issues like farm welfare, this can result in greater homogeneity in parties' policy positions in legislative settings. According to Aldrich and Rohde (2000), when issues such as farm animal welfare exhibit party congruence, members of the majority party will provide the leadership to secure desired policy change. Alongside recent research in the US (Shields, Shapiro, and Rowan 2017), the present findings suggest that traditional partisan differences on issues like farm welfare, may be shifting as parties adapt their policies in line with changing public opinion in an attempt to broaden their support base beyond their traditional constituencies.
(3) What are the principal characteristics of meso-electoral competition in relation to wildlife protection?
Key tropes in wildlife pledges across polities include the strengthening of legal protections (Scotland), habitat restoration (Wales), and species-specific interventions (Northern Ireland) (Table 4). Of all the animal welfare topics, the present study shows that the greatest Left-Right divergence is in relation to the sub-field of safeguarding wildlife habitats/ promoting biodiversity. Left parties predominate making 88 per cent of such pledges compared to 12 per cent by Right-of-centre parties (N = 434). We can relate this to the existing literature on partisan differences on environmental policy. For example, the current findings show that of the three main state-wide parties contesting the meso-elections, it is the Liberal Democrats that lead on the issue (17.8 per cent of pledges) (For example, 'We will ensure that EU Life+ funding is brought to Scotland and used for the enhancement of our most valuable and endangered species and habitats', (Scottish Liberal Democrats 2007, 17). This is consistent with Carter's (2006, 764) analysis of state-wide elections that found that since the early 1990s, 'it is the Centre-Left Liberal Democrats [that] have sought to present themselves as the greenest of the major parties, by consistently making the environment a core campaigning issue'. In contrast, Carter (2006, 763) found that, 'the Conservative and Labour parties have both pursued a strategy of preference-accommodation, characterized by a reactive approach to public opinion, events and issues, but resisting competition over the environment (although Labour has been consistently Table 4. Key tropes in wildlife pledges across polities.

Scotland -Strengthening of Legal Protections
'Increase protection for wildlife. Scottish Greens will support legislation that prosecutes those who harm or kill wildlife, and advocate that protection is extended to all wildlife, not only rare species' (Scottish Green Party 2016). 'To ensure legislation can be enforced, we will establish a unit within Police Scotland to specifically investigate wildlife crime' (Scottish Labour 2016). 'Tough new maximum penalties for those who commit crimes against wildlife' (Scottish National Party 2016) Wales -Habitat Restoration 'Support major projects to restore and create wildlife habitats across Wales' (Welsh Conservatives 2011, 8). 'We will launch a "Nature for People" programme of major habitat restoration and creation projects throughout Wales, with delivery coordinated between agencies, NGOs and landowners. This will contribute to achieving our biodiversity targets as well as delivering sustainable community development' (Plaid Cymru 2011, 16) 'We must protect our wildlife Biodiversity is under threat. Human activity means we are losing about 30 000 species per year. We have a historic opportunity not only to halt the degradation of our natural environment, but to begin to roll back two centuries of exploitation. We would: landscape conservation, using reform of planning system to encourage restoration of heath lands, woods, marshland and other important habitats' (Wales Green Party 2011, 18). Northern Ireland -Species -Specific Interventions 'Alliance advocates the Irish Hare be added to the list of animals that get full protection to combat its decreasing numbers' (Alliance Party 2011, 11). 'Introducing a Bee Strategy to address the drop-off in the number of bees, the consequence of which will be profound for the ecosystem' (Alliance Party 2016, 17). 'Promote green spaces and wildlife corridors in order to halt the loss of habitats and biodiversityprevent species decline, arrest the recent loss of wildlife by 2020' (Democratic Unionist Party 2011, 9) "greener" than the Conservative Party)'. This is consonant with the present analysis which shows the Labour Party made 10.6 per cent of wildlife protection pledges in the meso elections, compared to 5.8 per cent by the Conservatives. In light of Brexit, some Conservative pledges are striking. For example, 'Europe: [we will] work in closer co-operation with the European Commission to promote biodiversity' (Welsh Conservatives 2007, 6). However, Liberal Democrat and Labour pledges only offer a partial explanation of the Left's predominance in making wildlife pledges. The full answer lies in the significant level of attention to the issue by Left-of-centre civic nationalist parties and the Greens (who make 30.2 and 29.7 per cent of all meso wildlife pledges, respectively). When calculated as a proportion of each party's total animal welfare pledges, safeguarding protecting wildlife accounts for 48.5 per cent of the nationalist parties' pledges (compared to 24.8 per cent for the Greens). The latter finding resonates with recent research on partisan differences in environmental policy pledges in party manifestos in the UK. For example, Chaney's (2014b, 268) analysis revealed 'the greater propensity of nationalist parties to prioritise the environment'.
In turn, these partisan differences in the electoral discourse on wildlife policy are shaped by the actors and interest groups involved in this subfield and their relationship with different political parties. The key role of the Greens in advancing wildlife protection pledges in the mesoelections is easily explained because it is essentially a social movement turned political party (Barry and Doherty 2001) with intimate links to campaigning groups in civil society. Explaining the civic nationalist parties' leading role in the electoral discourse on wildlife policy also returns us to the Greens. There is an extended history of cooperation between the Greens and the SNP and Plaid Cymru. In each case this brought with it close links with environmental interest groups. Thus, in the Scottish Parliament the Greens signed a supply and confidence agreement with the SNP (c.2007). 10 In return, the SNP backed a climate change bill. Although it should be noted that more recently, the SNP's relationship with the Greens has not been so harmonious with the emergence of tensions with environmental interest groups (Kirkaldy 2018).
In the case of Plaid Cymru (that formed a coalition government with Welsh Labour in 2007-2011), there is also an extended history of formal alliance with the Green Party. Thus, a joint Plaid Cymru/ Green Member of Parliament was elected to the Westminster parliament in 1992. As in the Scottish case, one motive underpinning the nationalist-green alliance was to boost the two minority parties' share of the vote and likelihood of their candidates being elected. Yet, as Fowler and Jones (2006, 321) caution, there is nothing inevitable about such cooperation between the parties; rather 'relationships between environmentalism and nationalism shall remain a situationally contingent phenomenon'. As this suggests, electoral expediency is not the whole story behind such co-working. As Jane Dawson (2000, 23) notes, the intertwining of subgroup identity and environmentalism is not a new phenomenon; a special attachment to the land, flora, and fauna of a particular region has often played a role in nationalist writings and has frequently been used to reinforce mobilisation to nationalist or patriotic causes.
In a similar vein, Paul Hamilton (2002, 27) argues that 'nationalist movements will mine environmental discourse to enhance the legitimacy of their grievances and to widen their electoral appeal … civic nationalist goals can be complemented by environmental ones' (see also Fowler and Jones 2005;Lynch 1995). Such framing of policy pledges is evident in the current analysis of wildlife pledges. For example, 'A Plaid Cymru government will develop an ecosystem-based policy for oceans and coasts in consultation with scientists, stakeholders and environmental bodies. Its intention will be to: improve the health of our seas; protect marine wildlife' (Plaid Cymru 2003, 6) and, We will launch a 'Nature for People' programme of major habitat restoration and creation projects throughout Wales, with delivery coordinated between agencies, NGOs and landowners. This will contribute to achieving our biodiversity targets as well as delivering sustainable community development. (Plaid Cymru 2011, 7) In the case of Northern Ireland, earlier work (Chaney 2014a, 260) confirms a history of the two main nationalist parties' engagement with environmental policy-making: 'the nationalist SDLP accounted for the majority (44.9 per cent) of [manifesto] proposals on the environment 1999-2011. Second was Sinn Féin (27.9 per cent); thus combined, the nationalist parties account for almost three-quarters (73 per cent) of proposals'. As in the case of Scotland and Wales, Ó Broin (2015, 327) points to the nationalist parties' receptiveness to take on environmental interest groups' policy claims (as evidenced, for example, by the publication of the Sinn Féin policies: The Way Forward (Waste Management) 2001 and Towards Zero Waste 2004).

Conclusion
Hitherto in the UK and elsewhere, the impact of quasi-federalism or 'devolution' on the electoral politics of animal welfare has largely escaped scholarly attention. The foregoing analysis of over 1300 pledges on animal welfare in party manifestos for Scottish, Welsh and Northern Irish elections 1998-2017 reveals how devolution has introduced significant discontinuity in the process of animal welfare policy-making. Whether measured in absolute or relative terms, the present study shows a significant increase in the salience of animal welfare in party manifestos since meso elections were introduced in the UK in 1998/1999. This is significant for it reveals how meso-elections are providing new political spaces for electoral competition leading to new territorial approaches to animal welfare. In other words, it is leading to different 'sub-state' territorial animal welfare regimeswith contrasting government policies and law in the different polities of the UK. Electoral politics and the party politicization of animal welfare is the first stageor 'driver' of this process as parties compete by placing different animal welfare pledges before the electorates of Scotland, Wales and Northern Ireland.
The present study provides a further original insight into the meso party dynamics of animal welfare pledges. Whilst Garner's (1999, 92) pioneering work concluded that: 'that the most significant variable is party label, with concern for animals being associated with parties of the centre-Left'. The present analysis of partisan trends shows how this tends to over simplify matters. Whilst Left-dominance is borne out in the case of aggregate level analysis of animal welfare pledges (Research Question 1), the present analysis of farm animal welfare (Research Question 2) and wildlife protection (Research Question 3) also reveals distinctive partisan dynamics apply to different sub-fields. This supports earlier work (Lundmark, Berg, and Röcklingsberg 2018;Tosun 2017;Vogeler 2017) that underlines the need to examine party competition in sub-fields and shows how patterns and processes of party politicization are shaped by the actors and interest groups involved in the different subfields and their relation to different political parties.
In coming years the territorialization of animal welfare is likely to be subject to further change related to governance shiftsnotably, the post-Brexit 'repatriation' of European Union competence over aspects of animal welfare 11 and the public response to civic nationalist calls for Scottish and Welsh independence and Irish reunification. For now, in contrast to older notions of animals solely viewed as a resource and subject to limited policy pledges in state-wide Westminster election manifestos, the present analysis provides initial insight into a new 'sentient dynamic' in multi-level electoral politics in which party programmes are increasingly attuned to appeal to the symbiosis of humans and animals in distinctive ways in the different countries of the UK. However a caveat is necessary: political attention to animal welfare remains variable over electoral cycles and the political space for the devolved governments and parliaments to act on animal welfare may be constrained by future Westminster governments concerned to 'rein-in' devolved governance and themselves exercise powers repatriated from the EU, thus by-passing the devolved legislatures: tempus narrabo. Notes 1. Defined in terms of share of the popular vote. 2. Where necessary, hardcopy only versions of early manifestos were transcribed.
The software used was Nvivo 9 . 3. A worked example: 'we will legislate to introduce compulsory micro-chipping for dogs and cats'. This would be coded as two quasi-sentencesone under 'regulation', the other under 'pets'. 4. 4 instances 5. For example, in the present study, the word-length of Sinn Fein's manifestos over successive cycles is: 4234 (1998), 23 152 (2003), 19 097 (2007), 5601 (2011), 6896 (2016), and 5437 (2017). 6. As opposed to ratio measures of the number of pledges in relation to manifesto word length. 7. https://www.snp.org/ [Accessed 10 January 2020]. 8. For the purposes of cross polity comparison the 2017 Northern Ireland election is excluded. 35 animal welfare pledges were made in this election. This is lower than previous cycles and reflects the fact that generally party manifestos were shorter in length than previously owing to the somewhat hastily arranged nature of the ballot. Queen's Speech. In 2017 the UK Government said it believed the then draft Bill included provisions on animal cruelty within the National Assembly for Wales's (NAfW) legislative competence and that it intended to seek NAfW consent or the Bill, as required in law. Accessed 15 January 2020. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attach-ment_data/file/666576/draft-animal-welfare-bill-171212.pdf ORCID Paul Chaney http://orcid.org/0000-0002-2110-0436