Navigation – Plan du site

AccueilNuméros16DossierThe local in the Digital: compara...

Dossier

The local in the Digital: comparative insights from Cardiff and Rennes in the 2024 Elections

Le local et le numérique : comparaison entre Cardiff et Rennes lors des élections de 2024
Chiara Poletti et Marco Michieli

Résumés

Les élections législatives de 2024 au Royaume-Uni et en France offrent une occasion unique d'étudier, simultanément et de manière synchrone, comment les candidats mobilisent les liens territoriaux dans leurs campagnes numériques, dans différents contextes institutionnels. Cette étude compare la manière dont les candidats aux élections législatives de 2024 ont mis en œuvre le localisme sur la plateforme numérique X, à Cardiff (Pays de Galles), dans un système décentralisé, et à Rennes (Bretagne), dans un système centralisé. En nous appuyant sur la communication politique et les travaux de recherche sur la circulation médiatisée par les plateformes numériques, nous distinguons deux dimensions du localisme numérique : le répertoire (la manière dont les candidats tissent des liens locaux à travers la langue, les lieux, les enjeux et les ancrages matériels) et la circulation (les modalités d'engagement et d'amplification sur les plateformes). Grâce à une analyse mixte de 4 250 publications, nous examinons comment le localisme s'articule dans ces contextes institutionnels et si les modalités de circulation diffèrent en conséquence. Nos résultats montrent que les différences de répertoire correspondent aux arrangements institutionnels : les candidats de Cardiff déploient des répertoires locaux plus larges et transpartisanes, intégrant l'utilisation du gallois, les débats sur la décentralisation et des ancrages matériels tels que l'acier de Port Talbot et l'infrastructure du projet HS2. À Rennes, les candidats font un usage plus restreint et plus épisodique du localisme, concentré au sein des partis régionalistes. En revanche, les schémas de diffusion convergent dans les deux villes. Le contenu local suscite un engagement comparable dans les publications originales, mais est systématiquement éclipsé par les discours nationaux et les leaders des partis dans les réseaux de retweets. Cette déconnexion montre que les contextes institutionnels restent pertinents pour la manière dont les candidats intègrent le localisme, mais ces différences s'atténuent dans les schémas de diffusion.

Haut de page

Texte intégral

1. Introduction

1In France and the UK, institutional mechanisms to represent territorial interests within national parliaments are indirect, and the connection between candidates and their constituencies often becomes the key through which local concerns are introduced into the national political debate (Peinado García et al., 2025). Set against this background, the 2024 parliamentary elections in the UK and France provide a rare, synchronous vantage point to compare how ‘the local’ is mobilised by candidates across different institutional settings and incentive structures.

2This study builds on literature that values local ties in electoral politics and treats their salience as institutionally conditioned (Campbell and Cowley, 2014; Tavits, 2010; Zittel and Gschwend, 2008). It addresses gaps related to the prevalence of research on offline forms of localism as well as the narrow operationalisation of ‘the local’ and the limited attention to circulation patterns of localism online.

3Empirically, we analyse candidates’ activity on X (formerly Twitter) in a devolved (Cardiff) and centralised (Rennes) context. We draw on political communication and studies of engagement and amplification dynamics on social media platforms to interrogate in what forms and ways localism is mobilised by candidates on X and circulates within the platform. Our mixed-methods design combines descriptive distributions with an interpretive reading. We find a clear split: institutional differences are visible in repertoire but attenuate in circulation. On the repertoire side, Cardiff’s devolved context sustains a broader, cross-partisan and ‘thicker’ enactment of localism online, where attachments to places, material objects, institutional levers and minoritised language co-occur. In Rennes’ centralised setting, localism is not absent but appears narrower and more episodic, often concentrated in specific actors. On the circulation side, however, both cities see a recentring pattern: central frames attract more engagement than local ones; retweets disproportionately uplift central content and party leaders; and the upper tail of engagement is populated by posts that hook into broader, conflictual narratives. We present these findings as compatible with party-centred practices and circulation dynamics reported in prior work.

4The paper proceeds as follows. Section 2 situates the study within research on localism, digital campaigning and amplification dynamics on social media. Section 3 outlines the comparative context of Cardiff and Rennes. Section 4 details data, coding and limitations. Section 5 presents results and discusses implications for how localism is assembled differently across institutional arrangements and its digital circulation. Section 6 concludes.

2. Literature and analytical framework

5We structure the rationale behind this study in three steps. First, we examine how institutional contexts are associated with both the salience and the forms of localism in political communications. Second, we situate the study within the literature of political campaigns on social media, and the smaller research area of political localism online. Third, we clarify our theoretical approach, combining strategic communication perspectives with attention to how platform environments shape visibility.

2.1 The importance of being local, in different contexts

6Localism can be described as a repertoire of attributes and actions that link candidates to the constituencies they represent, such as birthplace, education, residence, and community involvement (Campbell and Cowley, 2014). This repertoire often serves as a heuristic for ‘behavioural localism’: the expectation that a candidate will act in alignment with community interests (Campbell et al., 2019), sometimes overriding the impact of other personal characteristics such as gender, religion, or education (Arzheimer and Evans, 2012; Schulte-Cloos and Bauer, 2023). Across Europe, localism is often treated as an entrenched feature of electoral politics, with voters valuing MPs from their constituency in national elections (Tavits, 2010). Candidates with deep local ties are seen as more trustworthy and legitimate, while ‘parachuted’ outsiders are frequently stigmatised (Hunt and Fontana, 2024). The relevance of local cues varies with institutional incentives and electoral system incentives (Zittel, 2009) and is often strategic and time sensitive, with increases around elections (Angenendt and Bukow, 2024; Silva et al., 2024).

7Although in the past some scholars downplayed the role of the ‘personal vote’ (Norris, 1995), subsequent work has shown that candidate-constituency connections remain a valuable strategic asset, particularly in competitive contexts (Vivyan and Wagner, 2016; Zittel et al., 2019) and in single-member systems that reward geographically concentrated support (Heitshusen et al., 2005; James and John, 2006; Pattie et al., 1995; Pattie and Johnston, 2003; Zittel and Gschwend, 2008). Appeals to localism become particularly salient in marginal constituencies and in those with strong regional identities, where candidates often position themselves as defenders of territorial interests (Arthur and Williams, 2019).

8The degree of institutional decentralisation significantly conditions how localism is articulated in electoral settings. In devolved systems such as the UK, research has found that local ties are increasingly emphasised, especially by Labour and devolved nations parties such as the Scottish National Party and Plaid Cymru (Cowley et al., 2024). In contrast, centralised systems like France’s tend to suppress the electoral salience of local identities, with national dynamics typically dominating voting behaviour (Jérôme and Lewis-Beck, 1999; Schrijver, 2004). Nevertheless, recent cross-national work indicates that local cues surface even in less ‘localist-friendly’ systems. Comparative work by Huwyler et al. (2025) notes that geographic representation appears even under proportional closed lists, suggesting that institutional discouragement reduces but does not eliminate local signalling (Fiva et al., 2020; Geese and Martínez-Cantó, 2022; Itzkovitch-Malka, 2021; Viganò, 2024). Indeed, in centralised settings like France, in-person strategies such as canvassing show the continuing importance of place-based engagement, particularly in culturally distinct regions, including Brittany (Cole et al., 2003; Pons, 2018). As such, localism has often been studied in its offline practices and only recently, with the rise of digital campaigning, research has begun to consider how candidates signal localness online (Huwyler et al., 2025).

2.2 Digital campaigning: decentralising affordances vs recentring forces

9Research has shown how social media have profoundly reshaped electoral campaigning, not only by transforming communicative formats but by altering the structural relations between candidates, voters, and parties (Bright et al., 2020; Jenkins et al., 2013). Social media’s decentralising tendencies make them an interesting place to look for localism. Social media can bypass gatekeepers, support citizen-led mobilisation and give visibility to local identities. Messages are less mediated than parliamentary behaviour and can enable more localised representation (Chadwick and Stromer-Galley, 2016; Lilleker et al., 2015). Moreover, social media can support decentralised participation by enabling citizen-led campaigns to assume functions historically managed by political parties, including fundraising and polling (Gibson, 2015). They can facilitate the engagement of less politically active groups, such as younger or first-time voters (Aldrich et al., 2015; Kimberlee, 2002) and allow politicians to engage in direct and informal communication with citizens (Enli and Rosenberg, 2018; Vliet et al., 2020). In doing so, social media can serve as spaces where territorial and cultural identities are reinforced, weaving together local narratives and political mobilisation (Arthur and Williams, 2019; Dipoppa and Grossman, 2020; Mason, 2018). However, research on localism online is still in its early stages. Despite the varied attributes and actions that constitute localism, online research has often operationalised it through narrow geographical references. For example, Huwyler et al. (2025) detects constituency references in recent studies of MPs’ social media in Germany, which built district-level dictionaries of municipality names to flag local content (e.g., tweets naming ‘Hamburg’, ‘Koblenz’, ‘Potsdam’) (Schürmann, 2023). Also, in their own study, they used a geospatial dictionary of toponyms and place names, while explicitly noting that this approach does not capture cues tied to people, organisations, events, linked content or images (Huwyler et al., 2025).

10Moreover, literature on localism online has paid limited attention to how local articulations intersect with recentring tendencies documented in broader digital campaigning research. Party discipline, leader-centred communication, and data-driven targeting consolidate central control, with campaign messaging on social media often replicating top-down structures (Lee, 2014; Grant et al., 2010). Larger parties frequently use social media to enforce centralised message discipline, limiting candidates’ autonomy over their digital communications (Swalwe et al. 2025). In the UK, only a minority of British MPs have been found to actively engage with voters on X, and even when they do, the interaction is often limited (Tromble, 2018). Similar patterns are visible in France, where retweet networks in electoral campaigns tend to coalesce around national political leaders, further reinforcing centralised party structures (Abdine et al., 2022). This dynamic is exacerbated by the increasing reliance on data-driven campaign techniques, including electoral databases and microtargeting, which shift authority from local operatives to centrally managed national teams (Denver et al., 2003, 2004; Römmele and Gibson, 2020). Together with algorithmic amplification (Gillespie, 2018; Marres, 2018), which may privilege polarising or generalised content (Jacobsen and Larsson, 2024; Kubin and von Sikorski, 2021; Marozzo and Bessi, 2017; Milli et al. 2025; Weismueller et al., 2022), decentralisation dynamics might be lost.

11Our study addresses this gap by expanding the operationalisation of ‘the local’ beyond geospatial mentions, examining both the repertoires candidates deploy and how these circulate within platform-mediated environments.

2.3 Localism: repertoire and circulation

12Our approach combines two perspectives. From political communication, we draw on established research that understands campaign messaging as intentional and strategic, oriented towards electoral gains (Peterson and Simonovits, 2018; Strömbäck and Kiousis, 2014). This approach is rooted in a tradition of framing analysis (Entman, 1993) and positions candidates as active agents in agenda-setting, selecting issues that resonate with targeted publics to mobilise support (Jacoby, 2000).

13We complement this with attention to how platform environments shape what becomes visible. Digital traces of localism represent candidates’ communication choices to link to people, objects and places of the constituency, but these connections are made visible through specific constraints of social media platforms (Gillespie, 2014). We note that user agency and resistance exist (Treré, 2018) but remain aware of the infrastructural conditions that configure what becomes visible and to whom (Møller Hartley et al., 2023; Marres, 2018). Platforms’ engagement metrics, retweet networks, and amplification dynamics selectively mediate which content gains traction (Milli et al., 2025). We therefore distinguish analytically between repertoire (i.e. the forms of localism candidates produce) and circulation (i.e. the patterns through which this content achieves visibility on the platform).

14We then turn our attention to specific cases of localism enacted online, in the context of Cardiff and Rennes.

3. Comparative context and case selection

15The pairing of Cardiff and Rennes offers a structured contrast in institutional conditions, devolved vs centralised, providing a lens to examine how localism is mobilised on social media across different governance frameworks. Here, we provide a concise overview of the similarities and contrasts in the two settings, building on prior comparative scholarship on Wales and Brittany (Cole, 2013).

3.1 Local identities and political systems

  • 1 Formal incorporation into centralised nation-states in the 16th century Brittany in 1532; Wales via (...)

16Brittany and Wales share deep-rooted cultural and linguistic identities with parallel histories of state-driven exclusion and suppression1. Both geographical regions are part of the Celtic cultural sphere and have retained Brittonic languages, Breton and Welsh, that have persisted despite centuries of marginalisation. Language has been a key vector of identity and cultural survival, with community-led initiatives such as Diwan schools in Brittany and Ysgolion Cymraeg in Wales playing a vital role in sustaining intergenerational language transmission (Wilson, 2023).

  • 2 Initiated with the 1998 Government of Wales Act and significantly expanded in 2006 and 2017.

17However, the recent political trajectories of these cultural movements have diverged markedly. In Wales, the process of devolution2 has led to a significant degree of institutional autonomy within the UK. Cardiff, the capital of Wales, operates within the devolved governance framework of the UK, where the Senedd (Welsh Parliament) holds legislative competences in areas including education, health, and language policy (Wilson, 2023). Although historically the political scene is dominated by the Welsh Labour Party, Plaid Cymru, the Welsh nationalist party, has actively leveraged devolution to promote autonomy-oriented issues (Arthur and Williams, 2019). This governance framework situates Welsh identity within formal political institutions, where it operates as a politically endorsed and legitimate feature of public life.

18In contrast, Rennes is embedded in France’s highly centralised political system. It remains a symbolic centre for Breton identity and culture, including efforts to preserve the Breton language. However, the absence of institutional regional autonomy continues to constrain the political impact of these initiatives (Cole and Williams, 2004), which depend largely on civil society and periodically encounter legal and political constraints (Cole et al. 2003; Cole and Baudewyns, 2004). This contrast between institutional support in Wales and centralisation in Brittany provides a critical lens to examine how digital performances of localism differ in the two systems and raises the question about how candidates perform localism in digital spaces where the symbolic and political weight of regional identity is not equally institutionalised.

3.2 Electoral systems

19The UK operates under a parliamentary system historically dominated by two parties, the Conservative and Labour, though recent decades have seen fragmentation and the rise of influence of regional parties such as Plaid Cymru and the Scottish National Party. France follows a semi-presidential model with a two-round parliamentary system that encourages a broader party field in the first round, followed by elite-driven alliance formation in the second (Gallagher and Mitchell, 2005; Parodi, 1978). These structural differences shape party dynamics and campaign strategies. In France, candidate selection and messaging are typically controlled by national leadership, with second-round alliances often negotiated by party elites in Paris and imposed locally (Bartolini, 1984; Gallagher et al., 2011). By contrast, the UK first-past-the-post (FPTP) system incentivises geographically concentrated support, making local campaigning central to electoral success, particularly for smaller and regional parties (Farrell, 1997).

20Neither country provides formal regional representation in their national parliaments. Constituency candidates serve as the primary, and often the only, channel through which local interests can be represented at the national level (Peinado García et al., 2025). This makes constituency-level politics particularly significant as a key site for articulating territorial identities and local concerns, even though research has shown that this connection often weakens or disappears once candidates reach central government (Peinado García et al., 2025). Taken together, these institutional and electoral variations lead us to expect differences in local signalling.

21With this study then, we wanted to expand the research on localism online by asking: how is online localism articulated across devolved (Cardiff/Wales) and centralised (Rennes/France) contexts? And are the forms of localism from different institutional contexts echoed in their circulation patterns online? This included examining different repertoires of objects, places, issues, and language that candidates mobilise in their tweets in different political contexts. Throughout the analysis, we then distinguish between repertoire (the presence of localism in posts) and circulation (patterns of platform engagement and retweet amplification).

4. Methodology

22This study combines quantitative analysis of post frequencies and platform engagement patterns with interpretative content coding to identify thematic patterns in the digital articulation of localism. Like Cole’s (2013) comparative research, this study adopts a mixed-methods approach, though on a more focused scale and within the specific context of digital campaigning. Whereas Cole’s analysis integrates institutional, policy, and survey data to trace regional capacity building over time, our interpretative lens is informed by political communication research and attention to platform-mediated visibility. We approach candidates’ social media activity as shaped both by strategic intent and by platform architectures, and in particular, the formats, engagement metrics, and amplification dynamics specific to X.

4.1 Data collection

23We examined candidates’ social media presence across major platforms (Facebook, Instagram, TikTok, YouTube, and X) and found that X was the only platform used consistently across all major parties and candidates in both Cardiff and Rennes (Figure 1). Recently, other research has used X to investigate constituency ties with candidates (Huwyler et al. 2025). Although its general audience reach is limited, it plays a key role in amplifying political messages through journalist and influencer networks (Russell, 2022; Taras and Davis, 2022). These features make X particularly relevant for analysing how candidates construct digital localism in a platform-centric campaign environment.

Figure 1: Social media platform usage by political party in Cardiff and Rennes (across all constituencies).

Figure 1: Social media platform usage by political party in Cardiff and Rennes (across all constituencies).
  • 3 Conservative and Unionist Party (Conservative), Green Party (Green), Heritage Party (Radical Right) (...)
  • 4 Ensemble - Horizons (Liberal), Ensemble Modem (Liberal), Ensemble - Renaissance (Liberal), Les Repu (...)

24Using web scraping tools, we collected posts from party-affiliated parliamentary candidates across the selected constituencies. Cardiff included 28 candidates3 across 4 constituencies; Rennes had 23 candidates4 across 4 constituencies, spanning one month before and one month after the elections in July 2024. However, the analysis focuses on the period between June 9 and July 10, 2024, to avoid overlap with the European elections in France, whose distinct voting system could distort local-national comparisons. The final dataset includes a total of 4,250 tweets (i.e., Rennes: 1879 and Cardiff: 2371), comprising original tweets and retweets from candidates, civil society actors, and news media. The dataset was manually validated for accuracy and completeness and contextual relevance, with additional information recorded on party affiliation, political position, and constituency. Candidate accounts were identified via party websites and cross-checked against official electoral lists; non-official/parody accounts were excluded, and where multiple accounts existed, the campaign-designated handle was retained.

4.2 Coding scheme and analysis

  • 5 Cole (2013) highlights that the concept of the ‘region’ is contested and politically charged, in bo (...)

25To analyse the tweets systematically, we developed a coding scheme structured around two core dimensions: localism repertoire and circulation in the broader online context. Within the localism dimension, we coded for Geographical focus, Type of content, and Language use in the tweets5 (see Table 1). We read the objects, places, issues, and language made visible in posts (or linked materials) as attachments that candidate use to signal local ties. We label attachments as strong when observed across multiple candidates/parties; otherwise, episodic.

26Within circulation in the broader online context, we coded Post source (for retweets) and Platform engagement. We relied on user engagement metrics (likes, retweets, replies) as indicators of post visibility. These metrics have been previously used to assess how posts ‘perform’ on platforms (Boulianne and Larsson, 2021; Gerbaudo et al. 2019; Farkas and Bene, 2021). For other research, user interactions represent a form of low-resource political mobilisation, as they require individuals to publicly associate with campaign content, carrying a degree of social risk by making political preferences visible (Bossetta and Schmøkel, 2023). However, following scholarship on platform-mediated visibility (Gillespie, 2014, 2018; Marres, 2021), we treat these metrics not as representations of coherent audiences but as platform-generated traces of interaction.

27Similarly to Gerbaudo et al. (2019), we operationalise platform engagement for each individual tweet as the sum of likes, retweets, and replies. This aggregated measure captures the total volume of interactions a post receives. To enable comparison across repertories and political contexts, we calculate engagement rates by dividing the sum of total engagement within each category (e.g., Local, Central, International) by the count of tweets in that category.

28We use three analytical strategies to examine circulation patterns. These strategies do not establish causal platform effects but allow us to assess whether observed distributions align with dynamics documented in prior research on digital campaign communication. First, retweet advantage. We compare engagement distributions for original posts versus retweets to examine whether visibility is reallocated through retweet networks and identify which actors benefit from this reallocation. Second, topic-engagement decoupling. We assess whether less frequent local topics achieve comparable or higher per-post engagement than more frequent ones, a pattern associated with skewed attention dynamics in platform-mediated environments (Marres, 2018, 2021). Third, polarisation hooks. We inspect top-engaged local posts for antagonistic or emotionally charged framings, which prior research identifies as systematically achieving higher visibility (Jacobsen and Larsson, 2024; Kubin and von Sikorski, 2021; Milli et al. 2025; Weismueller et al., 2022). The categories and codes are summarised in Table 1.

Table 1: Coding scheme description.

This table summarises the coding framework used to analyse candidates’ tweets. It includes three dimensions related to localism (geographical focus, type of content, and language use) and two dimensions that capture the broader communicative context: post source (for retweets) and platform engagement.

Localism dimension

Code

Description

Geographical focus

Local (0)

Referring to constituency or region.

Central (1)

Relating to central national political topics.

International (2)

Referring to global affairs.

Type of content

Policy-oriented (0)

Specific proposals or policy commentary.

Electoral Campaign (1)

Campaign actions or organisational messaging.

Personal (2)

Candidate’s personal life or content unrelated to policy/campaigning.

Language

Central (0)

English in Cardiff, French in Rennes.

Local (1)

Welsh (Cardiff) or Breton (Rennes).

Other (2)

Any other language (e.g. French candidates tweeting in English).

Broader context

Code

Description

Post source (for retweets)

Politician

Local, national, or EU-level party figures.

News Media

Journalists or news outlets.

Civil Society

NGOs, advocacy groups, community organisations.

Platform Engagement Rate

Sum of likes, retweets, and replies per tweet; we report this as a descriptive indicator of visibility for comparative purposes. For each tweet: sum of likes, retweets, and replies (Gerbaudo et al., 2019). For categorical comparison: sum of total engagement divided by tweet count within category, yielding average engagement per tweet.

29We conducted all coding collaboratively, with systematic cross-checking to ensure consistency and resolve ambiguities. Discrepancies were addressed through iterative discussion and refinement of the coding scheme. We coded tweets based on their dominant emphasis: in cases where ambiguity remained, we prioritised the element most explicitly identified through thematic prominence and visual cues, such as accompanying images or links.

4.3 Limitations

30While this study provides comparative insight into digital campaigning across Cardiff and Rennes, several limitations shape its scope and interpretation. Most notably, the analysis focuses exclusively on X. As noted, other platforms such as Facebook, Instagram, TikTok, and YouTube were not used consistently across parties or constituencies in either city. Even so, this platform-specific focus inevitably narrows the communicative landscape, privileging one digital environment while excluding others that may have attracted different audiences and communicative practices. While single-platform analyses have inherent limitations, this approach allows for systematic cross-party and cross-city comparison within a consistent methodological framework. As Boulianne and Larsson (2023) demonstrate, platform-specific affordances shape engagement patterns in ways that make cross-platform aggregation methodologically problematic without extensive calibration. Our X-focused design thus prioritises internal validity over breadth of platform coverage.

31We also recognise that the uptake of X varies between the two national contexts (i.e. X appears more peripheral in Rennes). As such, cross-city comparisons must be read with caution: observed patterns may stem as much from structural disparities in platform uptake as from candidate strategy.

32The absence of interview data or ethnographic observation places additional constraints on interpretation. Without direct insight into candidates’ intentions or audience responses, our analysis focuses on digital traces of localist performances rather than their deliberative or affective dimensions in real time.

33Moreover, while our analysis focuses on candidates’ communicative practices, we do not trace the formation of publics around these issues. When we consider localism repertoires and engagement rates, we do not assess whether these articulations successfully mobilise or sustain engaged publics. In this sense, our analysis does not claim to capture the full relational dynamics of digital political engagement.

34Finally, our analysis considers distributions of engagement and retweet networks but does not establish causal claims about platform mechanisms. Engagement patterns reflect multiple entangled dynamics, including platform affordances, party strategies, and audience composition, which cannot be isolated with our data. We present these patterns as descriptive traces that characterise the communicative environment within which localist repertoires operate, rather than as evidence of specific platform effects.

5. Results and discussion

35The comparative picture is marked by a split between repertoire and circulation. Repertoire mirrors institutional contexts: in Cardiff’s devolved setting, online localism appears broader, cross-partisan and more often; in Rennes’ centralised setting it is visible but narrower and more episodic. In contrast, circulation patterns converge across both cities despite these institutional differences. Part of this pattern is unsurprising given the national election contexts, where audience composition and news agendas are necessarily oriented toward central politics. Even then, and notwithstanding cross-country differences in X uptake noted above, the distributions display consistent features. The data show a local-over-central engagement advantage when looking at original posts and a pronounced retweet uplift that recentres attention on central content and party leaders, together with polarisation-sensitive tails.

5.1 Localism repertoire: broader in devolved settings, narrower in centralised ones.

36The distributions in Figure 2and Figure 3show that in Cardiff’s devolved setting the repertoire of localism cuts across content types (i.e. both policy and campaign content) and parties. Whereas in Rennes’ centralised setting it is mobilised mostly as campaign content and with limited cross-party reach, namely by the regionalist party Parti Breton and the Nouveau Front Populaire (NFP).

Figure 2: Distribution of local and central framing across tweet types (Cardiff vs Rennes).

Percentages by city and tweet content type; blue = local, red = central.

Figure 3: Local vs central framing in tweets by political party (Cardiff and Rennes)

Figure 3: Local vs central framing in tweets by political party (Cardiff and Rennes)

37Proportion of locally framed (blue) and centrally framed (orange) tweets by party within each city. Bars are normalised to 100% to enable comparison across parties with different posting volumes.

38Geospatial cues (Huwyler et al., 2025) appear often in the campaign register, where biographical proximity functions as the main local anchor used by candidates in both cities. The neighbourhood canvassing sequences, field trips, and logistical reminders convey belonging and familiarity with the places. At the same time, the frame of the ‘parachuted’ candidate acts as a negative signal of localism (as in Hunt and Fontana, 2024), emphasising the distance from the local electorate. The illustrations in Figure 4and Figure 5demonstrate this dual movement.

Figure 4: Examples of biographical proximity in candidate communication (Cardiff West and Rennes 3)6

Left: Plaid Cymru candidate foregrounding on-the-ground campaigning in named neighbourhoods, using canvassing imagery to signal familiarity and presence in the constituency. Right: NFP candidate combining school-based outreach with local interaction, performing proximity through everyday settings and face-to-face engagement. Together, they present a rhetorical-visual strategy that anchors political communication in lived local contexts close to the expectation of ‘behavioural localism’ (Campbell et al., 2019).

Figure 5. Examples of “parachute candidate” criticism (Cardiff West).7

The figure shows a Plaid Cymru candidate publicly criticising Labour’s selection as a “parachute” candidacy, framing the opponent as insufficiently locally embedded. The post mobilises proximity as a normative standard, positioning local residence, familiarity, and constituency presence as markers of legitimacy.

39Language usage (Figure 6 and Figure 7) on other hand, reflects institutional contexts. In Cardiff, Welsh is used across three parties. While in Rennes, Breton is exclusively used by the Parti Breton.

Figure 6: Use of Local language (Welsh and Breton) in candidates posts in Cardiff and Rennes.

Proportion of posts in local languages (Welsh in Cardiff; Breton in Rennes) versus central languages (English in Cardiff; French in Rennes). Blue bars = local languages; red bars = central languages.

Figure 7: Use of Local language (Welsh and Breton) in candidates posts to connect to local constituencies.8

Top: A Welsh-language post in which the candidate narrates doorstep engagement in a named neighbourhood and links it to locally salient issues (community spaces, access to sport, flooding), anchoring campaign communication in place-based concerns. Bottom: A Breton-language post that foregrounds Breton identity while connecting it to policy themes (housing, health, public services), using language choice as a marker of local belonging. Translation of Welsh text: “Yesterday morning we were in Caerau. A warm welcome as usual, plenty of conversations about the need for community spaces, equal access to sport, and concerns about flooding. An area that Labour has consistently let down.”

40Policy-focused posts also reflect differences in the political contexts. The frequencies in Table 2indicate that in Cardiff references to local policies are more frequent and diverse, while in Rennes they are fewer and more concentrated.

Table 2: Local policy coverage in Cardiff and Rennes

Shares of locally framed policy tweets by policy area within each city (percentages sum to 100% per city).

City

Topic

% of local-policy tweets

Examples

Cardiff

Autonomy and Local Governance

27.7

Devolution, criticism of Labour party, funds, electoral reform, national identity, and culture

 

Transportation and Infrastructure

18.1

HS2, transport access, parks, road safety, cycle lanes, infrastructure equity

 

Public and Healthcare services

14.6

NHS, housing, austerity, mental health, rent, social care, urban facilities

 

Employment and Professional Development

11.4

Tata Steel and Steel industries, employment precarity, local businesses, coal mining

 

Environmental issues

8.2

Pollution, green space, water management, climate,

 

Social Justice and Education

6.1

Food insecurity, free school meals, child poverty, elderly care, rural communities

 

Racial, Gender, and Minority Rights

3.8

LGBTQ+ rights, Pride, anti-fascism, freedom of expression

 

International politics

2.3

Gaza

 

Public Finance and Taxation

2.3

Anti-austerity, asking investments for Cardiff city

 

Economic Crisis and Purchasing Power

1.7

Budget, universal credit, taxes, tourism duties

 

Labour Policy

1.5

Workers’ rights, fair work, protests,

 

Immigration

1.2

Refugees

 

Law Enforcement

0.9

Devolution of power, Local policing, safer streets (i.e. street lighting)

 

Agriculture

0.3

Farmers

 

Total

100.0

 

City

Topics

% of local-policy tweets

Examples

Rennes

Autonomy and Local Governance

50.8

Autonomy, language rights, national identity, and culture

 

Public and Healthcare services

12.7

Housing and healthcare, welfare, rents

 

Transportation and Infrastructure

9.5

Stadium, cyclist accidents, roads (few specific localised frames)

 

Employment and Professional Development

7.9

Stellantis, job loss, railway sector

 

Labour Policy

6.3

Workplace accidents, strikes, job safety

 

Agriculture

4.8

Agriculture, general local rural concern

 

Environmental issues

3.2

Local enterprise, environment mentions

 

Immigration

1.6

Immigration mentioned in criticism of funds for refugees.

 

Economic Crisis and Purchasing Power

1.6

Local concerns are minimal

 

Social Justice and Education

1.6

Local concerns are minimal

 

Law Enforcement

0.0

Not present at local level

 

Total

100.0

 

41Across the political spectrum (Figure 8), Cardiff candidates reference devolution, local healthcare, housing, transport, environmental initiatives, minority rights, labour protections, and some international concerns in place-specific terms. In Rennes, mentions of local concerns are fewer, more episodic, and tied to individual parties (often Parti Breton).

Figure 8: Distribution of local policy mentions by party in Cardiff and Rennes

Cell values show the percentage of each party’s locally framed policy posts devoted to each policy area (columns sum to 100% within each party).

Cardiff local policies

Conservatives

Green party

Labour Party

Liberal Democrat

Plaid Cymru

Propel

Reform

Socialist Party

Autonomy and Local Governance

0%

27%

21%

42%

37%

25%

0%

0%

Transportation and Infrastructure

14%

4%

10%

33%

18%

33%

20%

0%

Public and Healthcare services

36%

31%

10%

17%

9%

9%

20%

20%

Employment and Professional Development

5%

4%

35%

0%

9%

3%

40%

60%

Environmental issues

9%

8%

4%

8%

4%

21%

0%

0%

Social Justice and Education

18%

8%

4%

0%

6%

4%

0%

0%

Racial, Gender, and Minority Rights

0%

2%

8%

0%

5%

1%

0%

0%

Public Finance and Taxation

9%

4%

0%

0%

3%

0%

0%

0%

International politics

0%

0%

0%

0%

5%

0%

0%

20%

Economic Crisis and Purchasing Power

5%

2%

2%

0%

1%

1%

0%

0%

Labour Policy

0%

6%

2%

0%

1%

0%

0%

0%

Immigration

0%

2%

0%

0%

1%

0%

20%

0%

Law Enforcement

5%

0%

2%

0%

1%

0%

0%

0%

Agriculture

0%

0%

0%

0%

0%

1%

0%

0%

Rennes local policies

Horizons

Lutte Ouvriere

MoDem

Nouveau Front Populaire

Parti Breton

Rassemblement national

Autonomy and Local Governance

0%

0%

3%

0%

97%

0%

Public and Healthcare services

0%

0%

0%

0%

100%

0%

Transportation and Infrastructure

0%

33%

0%

0%

67%

0%

Employment and Professional Development

0%

100%

0%

0%

0%

0%

Labour Policy

0%

25%

0%

0%

75%

0%

Agriculture

0%

0%

0%

33%

67%

0%

Environmental issues

50%

0%

0%

0%

50%

0%

Social Justice and Education

0%

0%

0%

100%

0%

0%

Immigration

0%

0%

0%

0%

0%

100%

42In Cardiff, Welsh devolution constitutes a credible institutional framework where local identity and policy delivery are aligned (Arthur and Williams, 2019). In Rennes, France’s centralised governance continues to limit the institutional weight of regional identity claims (Bartolini, 1984; Cole, 2013). Figure 9shows how ‘Autonomy and Local Governance’ in Cardiff is treated as an institutional matter, while in Rennes (Figure 10) autonomy is expressed primarily in a symbolic manner rather than connected to formal institutional pathways. References to Breton democracy, identity, or language preservation are used to articulate distance from French central power but remain loosely connected to formal structures of territorial governance. Across both examples, devolution is treated as an institutional reality, but it is mobilised to support competing claims about adequacy, future direction, and distributive justice.

Figure 9: Autonomy and local governance framing in Cardiff (Welsh Labour vs Plaid Cymru)9

Left: Welsh Labour framing devolution as an established and functional settlement that “delivers for Wales”, emphasising competence and continuity within devolved governance. Right: Plaid Cymru framing devolution as incomplete, calling for further devolution and fairer funding for Wales.

  • 10 Figure 10. URLs: https://x.com/AcademieDuGallo/status/1808872724325953563; https://twitter.com/Math (...)

Figure 10 : Autonomy and Local Governance in Rennes10

Posts that invoke autonomy and local governance in Rennes through symbolic and cultural registers rather than concrete institutional pathways.

43Material anchors further clarify these differences. In Cardiff, the steel plant of Port Talbot and the HS2 funding issue function as objects that bring together territorial identity, labour, and constitutional justice. The steel plant is mobilised as more than an economic site; it is a material anchor for performing Welsh identity (Figure 11). It ties political competition (and the preferability of the party) to the industrial legacy of Wales as well as to wider narratives about governance, territorial injustice, and constitutional arrangements (BBC News, 2024a; BBC News, 2024b). Similarly, the high speed HS2 railway funding debate is framed by candidates, particularly from Plaid Cymru, as an example of structural exclusion and symbolic marginalisation (BBC News, 2024c). The framing is explicitly territorial and material: candidates denounce the absence of HS2 rail infrastructure in Wales, highlighting how this physical disconnection makes visible wider political and constitutional asymmetries between Wales and Westminster (Figure 12). That is, they connect places and institutions to recognisable political stakes.

  • 11 Figure 11. URLs: https://x.com/WelshLabour/status/1802427582702641316 ; https://x.com/jevershed01/s (...)

Figure 11: The steel industry as a localism anchor in Cardiff11

Posts show how Welsh steel, especially Port Talbot, functions as a shared localism anchor in candidate communication. Left: Labour post foregrounds steel as a defining feature of Welsh communities and presents protection of the industry as a commitment to local livelihoods. Right: A critical post mobilises the same reference to argue against Labour, making the steel industry the stake of who has local interests at heart.

  • 12 Figure 12. URLs: https://x.com/Plaid_Cymru/status/1808870561566593405

Figure 12. HS2 as anchor for local concerns in Cardiff.12

The figure shows a Plaid Cymru post mobilising HS2 as a proxy for Welsh local interests, despite the project’s physical absence from Wales. The tweet frames HS2 funding as a territorial injustice, positioning the campaign as a demand for financial redress and constitutional fairness against the UK government.

44In Rennes, candidates also address local industrial concerns, notably the protests surrounding Stellantis’s announced workshop closure at La Janais, an automotive production site located in Chartres-de-Bretagne, on the outskirts of Rennes (France 3 Régions, 2024; Ouest-France, 2024). However, these local issues tend to be narrated in broader terms of labour conflict and global capitalism, with a weaker connection to the identity of the territory (Figure 13).

  • 13 Figure 11. URLs: https://x.com/cgtinterim/status/1808470119565115484

Figure 13: Local focus and policy framing in Rennes: Stellantis as a labour issue13

Posts about Stellantis in Rennes foreground local economic disruption while framing it primarily through class and political-economic critique rather than territorial identity. Job losses, subcontracting, and workplace insecurity are narrated as consequences of corporate strategy and capitalist power, placing emphasis on labour struggle and social conflict rather than regional autonomy or cultural distinctiveness.

45Finally, in both contexts, hyper-local symbols: parks, speed limits, bus lines, cyclist safety, construction sites, operate as points of contact that translate politics into the ordinary (Figure 14). Across the examples, political critique is anchored to tangible changes in a familiar public space (e.g., trees, pitches, facilities), translating abstract claims about governance or representation into materially observable local impacts and everyday concerns.

  • 14 Figure 14. URLs: https://x.com/rodneyberman/status/1803687539984769081; https://x.com/neiljmcevoy/s (...)

Figure 14: Hyper-local references in Cardiff14

The figure shows how candidates in Cardiff perform localism through hyper-local place references, using a public park as a shared point of recognition.

46In Rennes, similar hyper-local references do appear, but less frequently. Posts about cyclist safety, stadium development, and local road conditions however demonstrate some anchoring in everyday spaces (Figure 15). These examples also demonstrate how candidates assemble localism through situated symbols and local concerns. Everyday infrastructures such as steel factories, transport networks, public parks, and speed limits anchor political discourse in the lived spaces and institutional memories of the constituency and can mobilise public engagement. In this sense, localist articulations online could serve as entry points for more grounded, community-specific concerns and potentially broaden the range of actors and issues involved in digital campaigning.

Figure 15: Hyper-local references in Rennes15

The figure shows candidates in Rennes drawing on hyper-local issues, such as stadium development and roadworks, to situate political communication in recognisable, everyday concerns. Across the examples, the local is mobilised through references to familiar sites and disruptions that residents encounter directly, providing concrete anchors for critique, responsibility claims, and campaign positioning.

47Overall, we found in our data that localism on X is built by candidates through the combination of several elements, that move beyond the mention of geographical location (Huwyler et al., 2025). In Cardiff, Welsh language cues, concrete material stakes (e.g., steel, HS2 funding), the institutional vocabulary of devolution, and everyday service issues, makes localism a layered political claim. In Rennes, by contrast, localism is expressed through a smaller repertoire, most visibly Breton language and autonomy symbolism concentrated in specific parties, so it functions more as a party-linked identity marker than as a cross-cutting frame that travels across policy domains.

48However, as Section 5.2 will demonstrate, these repertoire differences do not automatically translate into circulation advantages.

5.2 Circulation traces - Institutional effects attenuate in platform-mediated environments

49Section 5.1 showed that candidates’ repertoires reflect their institutional environments: localism is broader and more consistently articulated in Cardiff, whereas in Rennes it appears more episodic. We now examine how these repertoires circulate through engagement patterns. Our analysis suggests that circulation dynamics shape which local content achieves visibility in ways that may not straightforwardly reflect institutional contexts. We show that retweet networks systematically privilege central content, and that within local content, posts combining territorial identity with grievances or emotionally resonant framings tend to generate higher engagement than those focused primarily on policy detail.

50Candidates in both cities produced similar proportions of original content and retweets: 35% original posts and 65% retweets. When examining original tweets (Figure 16), posts with local focus receive higher engagement than those referring to central political contexts in both cities.

Figure 16: Engagement rate for candidates’ original posts in Cardiff/Rennes

Engagement rates for candidates’ original posts in Cardiff and Rennes, distinguishing locally framed (blue) and centrally framed (red) content.

51Once we consider retweets (Figure 17), the engagement pattern shifts. Retweets generate far higher engagement than original posts, and this amplification disproportionately benefits central content. In Cardiff, retweeted central content accumulates 22 times higher engagement than retweeted local posts. In Rennes, retweeted central content receives 9 times more engagement on average than local retweets.

Figure 17: Engagement rate by post type: original posts vs retweets in Cardiff/Rennes

Engagement rates for original posts and retweets in Cardiff and Rennes, distinguishing locally framed (blue) and centrally framed (red) content.

52Prior research identifies several factors that may contribute to this pattern. The hierarchical structure of party communication networks represents one such mechanism (Swalwe et al., 2025). Examining the most frequently retweeted accounts reveals that candidates overwhelmingly recirculate political actors’ voices rather than diversifying visibility to include local media, civil society organisations, or community groups (Figure 18). In Cardiff, politicians account for approximately 80% of retweeted sources; in Rennes, 83%. Media and civil society together capture less than 20% in both cities. This concentration creates an intra-political circulation circuit where visibility remains concentrated among party actors rather than diversifying toward civic organisations, local media, or community voices.

Figure 18: Retweeted actor types (Cardiff and Rennes)

Distribution of retweeted content by actor type in Cardiff and Rennes. Percentages are calculated within each city.

53Table 3 reports the accounts most frequently retweeted by candidates in Cardiff and Rennes, revealing different coordination logics that nonetheless converge on a similar outcome: visibility concentrates around a small set of high-status partisan accounts.

Table 3: Top 10 most retweeted political accounts in Cardiff and Rennes

Top ten politician accounts most frequently retweeted by candidates in Cardiff and Rennes. N = volume of retweets; % Local = share of locally framed content in the retweeted material; Eng Rate = average engagement per retweet.

City

Rank

Account

Account Type

N Retweets

% Local

Eng Ratio

Cardiff

1

Plaid Cymru (@Plaid_Cymru)

Welsh nationalist party advocating for Welsh independence

116

78.4

405

 

2

Keir Starmer (@Keir_Starmer)

Official account of the UK Labour Party leader

97

1.0

4,513

 

3

UK Labour (@UKLabour)

Official account of the Labour Party

41

4.9

35,528

 

4

Anthony Slaughter (@as_penarth)

Local leader (Welsh Green Party Leader)

36

55.6

240

 

5

Phil Davis (@phildavies1995)

Local leader (Welsh Green Party Leader)

34

70.6

89

 

6

Welsh Labour (@WelshLabour)

Welsh section of the Labour Party

33

54.5

339

 

7

Rhun ap Iorwerth (@RhunapIorwerth)

Leader of Plaid Cymru

19

94.7

810

 

8

The Green Party (@TheGreenParty)

Official account of the UK Green Party

18

5.6

3,042

 

9

Mabon ap Gwynfor (@mabonapgwynfor)

Member of the Senedd (Welsh Parliament) for Plaid Cymru

18

77.8

226

 

10

Jo Stevens (@JoStevensLabour)

Local candidate (Labour for Cardiff East)

15

73.3

86

City

Rank

Account

Account Type

N Retweets

% Local

Eng Ratio

Rennes

1

Nathalie Arthaud (@n_arthaud)

National spokesperson for Lutte Ouvrière

67

0.0

1,291

 

2

Parti Breton (@PartiBreton)

Official account of the Breton Party

50

90.0

348

 

3

Lutte Ouvrière (@LutteOuvriere)

Official account of LO

50

0.0

72

 

4

Avec Tristan (@AvecTristan)

Local candidate (NFP)

35

88.6

69

 

5

Hubert Ludo (@hubert_ludo)

Local candidate (PCF)

24

25.0

37

 

6

Chantal Gomez (@chantalgomez38)

Party member (Lutte Ouvrière)

23

0.0

32

 

7

Marie Ducruet (@MarieDucruet_LO)

Party member (Lutte Ouvrière)

18

0.0

62

 

8

Stéphane Pellegrini (@StephPellegrini)

Local candidate (Lutte Ouvrière)

17

0.0

18

 

9

Jean-Pierre Mercier (@JPierre_Mercier)

Spokesperson Lutte Ouvrière

17

0.0

110

 

10

Tobeilem Gregory (@lo_tobeilem)

Local candidate (Lutte Ouvrière)

13

0.0

31

54In Cardiff, the top retweeted accounts mix devolved-national and UK-wide actors. Alongside Plaid Cymru and Welsh political figures, national party elites and official party accounts (e.g., Keir Starmer; UK Labour; The Green Party) feature prominently, consistent with “extended party network” dynamics (Yoon et al. 2022) or similar situations in which lower-tier actors amplify elite messaging to expand reach (Swalwe et al., 2025). However, the content and attention profiles diverge sharply. National accounts circulate overwhelmingly non-local framing (e.g., Keir Starmer: 1.0% local; UK Labour: 4.9% local) but attract far higher engagement: UK Labour retweets average 35,528 engagements, versus 405 for Plaid Cymru (78.4% local), an approx. 88-times difference despite Plaid receiving more retweets overall (116 vs 41). This disparity aligns with evidence that “authority” cues increase sharing, making established accounts structurally advantaged in diffusion (Weismueller et al., 2022).

55In Rennes, by contrast, retweeting is dominated by a closed organisational circuit: 7 of the top 10 accounts are Lutte Ouvrière (LO)-affiliated, indicating dense intra-party reinforcement rather than cross-party or regionally anchored circulation. This maps onto retweet-network findings from the 2022 French presidential election, where communities cluster around candidates/parties and interactions are strongly modular (Abdine et al., 2022). Here, LO-linked accounts also circulate overwhelmingly non-local framing (often 0% local), with only limited counterweight from explicitly Breton-oriented actors (e.g., Parti Breton: 90% local).

56These coordination patterns operate within platform-mediated visibility regimes, where circulation is shaped not only by political relevance or campaign strategy but also by ranking logics and engagement metrics that systematically advantage already-prominent sources (Marres, 2018). This matters because engagement-optimised ranking can amplify divisive content and may underperform users’ stated preferences, meaning that what travels furthest is not necessarily what users say they want to see (Milli et al., 2025).

57Accordingly, in our data localism is structurally disadvantaged in circulation terms: higher posting volume does not translate into proportionally higher engagement. This is especially visible when we restrict attention to locally framed policy posts.

58Figure 20(Cardiff) and Figure 21(Rennes) plot for each policy area, the number of locally framed policy tweets (x-axis, N) against the mean engagements per tweet (y-axis, engagement rate). If engagement closely tracked topic volume, we would expect a broadly upward pattern in which higher N corresponds to higher engagement. Instead, topics with similar volumes attract very different engagement rates, and some low-volume topics match or exceed high-volume topics. We refer to this as topic-engagement decoupling in the straightforward sense that variation in engagement is not explained by variation in posting frequency.

Figure 20: Engagement with locally framed policy posts in Cardiff, by policy area

Scatter plot of locally framed policy-oriented posts in Cardiff, disaggregated by policy area. The x-axis reports the number of posts (N) in each policy area; the y-axis reports the mean engagement per post (engagement rate). Points represent policy areas; dispersion indicates that similar posting volumes can correspond to very different engagement levels.

Figure 21: Engagement with locally framed policy posts in Rennes, by policy area

Scatter plot of locally framed policy-oriented posts in Rennes, disaggregated by policy area. The x-axis reports the number of posts (N) in each policy area; the y-axis reports the mean engagement per post (engagement rate). Points represent policy areas; dispersion indicates that similar posting volumes can correspond to very different engagement levels (note: presence of outlier “Immigration”).

59A closer reading suggests that local-policy posts tend to travel further when they condense localism into sharper conflict cues (e.g. institutional contestation, accountability claims, boundary-making, grievance).

60In Cardiff, Autonomy and Local Governance is both frequent and moderately engaging (N = 95; ER = 251.67; see Figure 9for how devolution is mobilised as institutional matter with competing claims about adequacy and fairness). However, comparatively lower-volume domains can outperform it, including Law Enforcement (N = 3; ER = 259.33) and Employment and Professional Development (N = 39; ER = 266.87). These higher-engagement domains are also those where localism is articulated through material or constitutional contestation (e.g., Tata Steel and local industrial futures, Figure 11; HS2 as proxy for Welsh exclusion and constitutional asymmetries, Figure 12), often framed as grievances against UK-level priorities (see Section 5.1).

61In Rennes, the decoupling is more extreme because engagement concentrates in a small number of conflictual cues. While Autonomy and Local Governance appears relatively often (N = 32, predominantly articulated through symbolic and cultural registers by Parti Breton rather than concrete institutional pathways; Figure 10), the highest average engagement comes from Labour Policy (N = 4; ER = 615.50, including Stellantis job losses framed through class and political-economic critique; Figure 13) and a single Immigration post (N = 1; ER = 5,203, Figure 22). By contrast, other local-policy domains such as Transportation and Infrastructure (N = 6; ER = 15.17, topics like stadium development and road conditions documented in Figure 15) and Environmental issues (N = 2; ER = 8) attract comparatively weak interaction.

62This Immigration post outlier illustrates the starkest expression of topic-engagement decoupling. The RN Bretagne (Figure 22) assembles a nationally salient topic (immigration) with regional budgets and territorial grievance through out-group hostility. This illustrates an emotion-over-argument configuration in which the post’s visibility is driven less by policy detail than by grievance and boundary-making. With 5,203 engagements, it demonstrates how divisive issue framing can function as an attention-generating anchor, operating through an emotion-over-argument configuration in which visibility is driven by boundary-making and grievance rather than policy discussion.

  • 16 Figure 18. URLs: https://x.com/RNBretagne/status/1810348109261885872

Figure 22: High-engagement local-policy retweet in Rennes: RN framing on immigration16

The figure shows a retweeted RN Bretagne post that generated the highest engagement among locally framed policy retweets in Rennes. The post centres immigration through a fiscal grievance frame, criticising regional spending on migrant rescue and casting it as a burden on local taxpayers. Its outsized engagement illustrates how highly divisive issues can function as attention-generating anchors within local-policy communication, attracting disproportionate interaction compared to less polarising local topics.

63In Cardiff, a comparable dynamic appears in a post that achieved the highest engagement among all locally framed content (N = 1; ER = 6,117; Figure 23). Rather than policy discussion or campaign messaging, the post mobilises Welsh identity through grievance, linking territorial belonging to perceived neglect or constitutional unfairness, echoing the institutional and identity framings documented in Section 5.1 but stripped of any actual policy implication. This supports the broader pattern that contentious, emotionally resonant local grievances can travel further than routine local-policy content.

Figure 23: High-engagement local post in Cardiff: identity-linked grievance framing.17

The figure shows a locally framed post that attracted among the highest engagement levels in Cardiff. Rather than offering detailed policy discussion, the post mobilises Welsh identity through a grievance frame, linking local belonging to perceived neglect or unfairness (e.g., recognition, redistribution, devolution-related discontent). This mirrors the broader pattern in which contentious, emotionally resonant local grievances travel further than routine local-policy content, generating disproportionate interaction.

64Across both cities, these distributions align with comparative evidence that policy posts tend to generate lower engagement than other post types on social media platforms (Boulianne & Larsson, 2021). Several mechanisms help explain this pattern. First, policy content focused on administrative detail or service delivery demands more cognitive processing, whereas platforms reward rapid affective responses that generate quick engagement signals. Weismueller et al. (2022), analysing over 10,000 political tweets, show that content where emotions outweigh argument quality is more likely to be shared, and that authority (including follower count) increases shares; they also find that the advantage of positive emotional content weakens among ideologically extreme users. Second, platform architectures themselves shape these dynamics: algorithmic audit evidence indicates that engagement-based ranking can amplify divisive content relative to alternative ranking principles (Milli et al., 2025).

65This does not render local repertoires irrelevant: candidates continue to articulate local concerns, and locally framed original posts can achieve meaningful engagement. However, once content enters wider circulation through retweet networks, platform-amplified asymmetries and authority effects become more decisive, and the locally framed messages that attain high engagement tend to be those most legible to platform visibility logics, namely, those that generate fast reactions rather than sustained policy deliberation.

6. Conclusion

66This article examined how online localism is assembled under different institutional arrangements and how these repertoires circulate on social media. Drawing on political communication and attention to platform-mediated visibility, we analysed candidates’ posts on X during the 2024 election campaigns in Cardiff (devolved) and Rennes (centralised). Conceptually, the study advances two key contributions. First, it moves from asking whether candidates “mention places” to tracing how the local is enacted as a repertoire of attachments (objects, places, issues, language) under specific institutional conditions. This helps to capture forms of localism that standard geographical proxies might miss. Second, it analytically separates repertoire formation from circulation. The analytical strategies we apply (i.e. retweet advantage, topic-engagement decoupling, and polarisation hooks) can help contextualise thicker repertoires of localism within broader patterns of circulation online. Analytically, we treat platforms as mediators rather than neutral channels, recognising that the visibility of localist content is shaped not only by candidate strategy but also by engagement dynamics, retweet networks, and platform architectures. Substantively, the comparison reveals a critical split. Repertoire formation aligns with institutional contexts: in Cardiff’s devolved setting, localism displays broader, cross-partisan assemblages with recognisable institutional levers (devolution, Welsh language, material anchors); in Rennes’ centralised setting, repertoires appear narrower and more episodic. However, circulation patterns attenuate these institutional differences: in both cities, engagement skews toward national content, is redistributed upward through retweet networks and clusters around politician-heavy sources and leader cascades.

67Our findings complicate early optimism about social media as inherently decentralising forces in political participation (Aldrich et al., 2015; Gibson, 2015; Enli and Rosenberg, 2018). That optimism often assumed that lowered production costs and direct-to-audience communication would naturally redistribute visibility toward constituency-level content and peripheral actors (Kimberlee, 2002). Our data challenge this assumption in a specific sense: while candidates do produce substantial local content, and such content can achieve engagement within original posts, its visibility is systematically diminished once it enters circulation through retweet networks. Local repertoires are structurally disadvantaged not at the point of production but at the point of amplification, where hierarchical party structures, authority effects, and engagement-optimised ranking converge to privilege national frames and established accounts. This does not mean social media offers no advantages for local content relative to pre-platform environments. Candidates can now articulate constituency concerns directly without traditional gatekeepers, and locally framed original posts do generate interaction. However, the redistribution of visibility through platform-mediated circulation introduces new forms of asymmetry that may offset these gains. Whether the net effect represents an improvement relative to offline campaigning remains difficult to assess without direct comparison.

68This study has several limitations. It focuses on a single platform (X), a restricted time window, and two cities with different levels of platform uptake; it lacks follower/impression baselines and audience or campaign-side interviews. Our interpretations are explicitly non-causal and stop at compatibility with documented practices. Future research could extend the design across platforms and election cycles, integrate audience or network-structure data and pair it with qualitative or experimental work on campaign decision-making and reception. Comparative work examining the same constituencies across offline and online channels, or across pre- and post-platform election cycles, would help establish whether observed circulation asymmetries represent new constraints or continuities with existing power structures. Finally, the findings open an agenda on where localism might travel online. Our analysis points to the importance of alternative channels of visibility (e.g. civil-society actors, locality-specific media, or format choices that foreground place, object, and language attachments) capable of sustaining local content beyond initial posting. Whether such devices can counteract the recentring dynamics we document remains an empirical question; what our evidence shows is that localism can be assembled richly in candidates’ repertoires, but its circulation on X is selectively mediated.

Haut de page

Bibliographie

Abdine Hadi, Guo Yanzhu, Rennard Virgile, & Vazirgiannis Michalis (2022). ‘Political communities on X: case study of the 2022 French presidential election’. arXiv preprint. https://arxiv.org/pdf/2204.07436 .

Aldrich John H., Gibson Rachel K., Cantijoch Marta & Konitzer Tobias (2015). Getting out the vote in the social media era: Are digital tools changing the extent, nature and impact of party contacting in elections? Party Politics, 22(2), pp. 165-178. https://doi.org/10.1177/1354068815605304.

Angenendt Michael & Bukow Sebastian U. (2024). Strategic choices and electoral gains: Unpacking constituency campaign effectiveness in Germany’s mixed-member electoral system. Journal of Elections, Public Opinion and Parties. https://doi.org/10.1080/17457289.2024.2337437

Arthur Rudy & Williams Hywell T. (2019). The human geography of X: Quantifying regional identity and inter-region communication in England and Wales. PloS one, 14(4), p.e0214466. https://doi.org/10.1371/journal.pone.0214466

Arzheimer Kai & Evans Jocelyn (2012). Geolocation and voting: Candidate–voter distance effects on party choice in the 2010 UK general election in England. Political Geography, 31(5), pp. 301–310. https://doi.org/10.1016/j.polgeo.2012.04.006 .

Bartolini, Stefano (1984). Institutional Constraints and Party Competition in the French Party System. West European Politics, 7/4, pp. 103–27. https://doi.org/10.1080/01402388408424499

BBC News (2024a). Steel jobs secured with government support. [online]. https://www.bbc.co.uk/news/articles/ckvvk1dy99qo

BBC News (2024b). The UK government and Tata agree deal to secure jobs. [online]. Available at: https://www.bbc.co.uk/news/uk-wales-66830867

BBC News (2024c). Wales should get £4bn HS2 compensation - Plaid Cymru. [online]. Available at: https://www.bbc.co.uk/news/articles/c3ggw27403lo

Bossetta Michael & Schmøkel Rasmus (2023). Cross-platform emotions and audience engagement in social media political campaigning: Comparing candidates’ Facebook and Instagram images in the 2020 US election. Political Communication 40.148-68. https://doi.org/10.1080/10584609.2022.2128949

Boulianne Shelley and Larsson Anders Olof (2021). Engagement with candidate posts on Twitter, Instagram, and Facebook during the 2019 election. New Media and Society 25(1), pp. 119-140. https://doi.org/10.1177/14614448211009504

Bright Jonathan, Hale Scott, Ganesh Bharath, Bulovsky Andrew, Margetts Helen & Howard Phil (2020). Does Campaigning on Social Media Make a Difference? Evidence From Candidate Use of Twitter During the 2015 and 2017 U.K. Elections. Communication Research, 47(7), pp. 988-1009. https://doi.org/10.1177/0093650219872394

Campbell Rosie & Cowley Philip (2014). What voters want: Reactions to candidate characteristics in a survey experiment. Political studies, 62(4), pp.745-765. https://doi.org/10.1111/1467-9248.12048

Campbell Rosie, Cowley Philip, Vivyan Nick & Wagner, Markus (2019). Why friends and neighbors? Explaining the electoral appeal of local roots. The Journal of Politics, 81(3), pp. 937-951. https://doi.org/10.1086/703131

Chadwick Andrew & Stromer-Galley Jennifer (2016). Digital Media, Power, and Democracy in Parties and Election Campaigns: Party Decline or Party Renewal? The International Journal of Press/Politics, 21(3), 283-293. https://doi.org/10.1177/1940161216646731

Cole Alistair (2013). Political institutions, public and elite opinions in Wales and Brittany. In Beyond devolution and decentralisation. Manchester University Press. https://doi.org/10.7765/9781847792105.00010

Cole Alistair & Baudewyns Pierre (2004). Devolution, Decentralisation and Public Opinion in Wales and Brittany. Comparative European Politics, 2, pp.73-96.

Cole Alistair & Williams Colin (2004). Institutions, identities, and lesser-used languages in Wales and Brittany. Regional and Federal Studies, 14(4), pp.554-579. 10.1080/1359756042000315487

Cole Alistair, Jones J. Barry, Williams Colin, Loughlin John & Storer Alan (2003). Political

institutions, policy preferences and public opinion in Wales and Brittany. Contemporary Wales 15 (1) pp. 89-110

Cowley Philip, Gandy Robert J. & Foster Scott (2024). Increasingly local: The regional roots of British members of parliament, 2010–2019. The Journal of Legislative Studies, 30(3), pp.253-264.

Denver David, Hands Gordon, Fisher Justin & MacAllister Iain (2003). Constituency Campaigning in Britain 1992-2001: Centralisation and Modernisation. Party Politics, 9(5), pp. 541-559. https://doi.org/10.1177/13540688030095002 .

Denver David, Hands Gordon & MacAllister Iain (2004) The Electoral Impact of Constituency Campaigning in Britain, 1992–2001. Political Studies, 52(2), pp. 289–306. https://doi.org/10.1111/j.1467-9248.2004.00480.x.

Dipoppa Gemma & Grossman Guy (2020). The effect of election proximity on government responsiveness and citizens’ participation: Evidence from English local elections. Comparative Political Studies 53(14), pp. 2183-2212. https://doi.org/10.1177/0010414020912290

Enli Gunn & Rosenberg Linda T. (2018). Trust in the age of social media: Populist politicians seem more authentic. Social Media + Society, 4(1), pp. 1–11. https://doi.org/10.1177/2056305118764430

Entman Robert M. (1993). Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication 43 (4): 51–58. https://doi.org/10.1111/j.1460-2466.1993.tb01304.x

Farrell David M. (1997). Comparing electoral systems. Macmillan Press Ltd

France 3 Régions (2024). CEAT de Longvic : les ex-salariés empêchent Samsung de récupérer 8 000 téléphones. [online] Available at: https://france3-regions.francetvinfo.fr/bourgogne-franche-comte/cote-d-or/ceat-de-longvic-les-ex-salaries-empechent-samsung-de-recuperer-8000-telephones-2834330.html

Farkas Xénia, & Bene Márton (2021). Images, politicians, and social media: Patterns and effects of politicians’ image-based political communication strategies on social media. The International Journal of Press/Politics, 26(1), pp. 119–142. https://doi.org/10.1177/19401612209595

Fiva Jon H., Halse Askill. H., & Smith Daniel. M. (2020). Local representation and voter mobilisation in closed-list proportional representation systems. SSRN. https://doi.org/10.2139/ssrn.3447662

Gallagher Michael & Mitchell Paul (eds). (2005). Introduction to Electoral Systems, In The politics of electoral systems. Oxford University Press. https://doi.org/10.1093/0199257566.001.0001

Gallagher Michael, Laver Michael & Mair Peter (2011). Representative Government in Modern Europe, 5th revised edition, London, McGraw-Hill, - https://hdl.handle.net/1814/16654

Geese Lucas & Martínez-Cantó Javier (2022). Working as a team: Do legislators coordinate their geographic representation efforts in party-centred environments? Party Politics, 29(5), pp. 918-928. https://doi.org/10.1177/13540688221098157

Gerbaudo Paolo, Marogna Federico & Alzetta Chiara (2019). When “positive posting” attracts voters: User engagement and emotions in the 2017 UK election campaign on Facebook. Social Media+ Society 5.4: 2056305119881695.

Gibson Rachel K. (2015). Party change, social media and the rise of ‘citizen-initiated’ campaigning. Party Politics, 21(2), pp. 183–197. https://doi.org/10.1177/1354068812472575.

Gillespie Tarleton, (2014). The relevance of algorithms. In T. Gillespie, P. J. Boczkowski, and K. A. Foot (Eds.), Media Technologies: Essays on Communication, Materiality, and Society. MIT Press.

Gillespie Tarleton (2018). Custodians of the Internet: Platforms, content moderation, and the hidden decisions that shape social media. Yale University Press.

Grant Will J., Moon Brenda & Busby Grant Janie. (2010). Digital dialogue? Australian politicians’ use of the social network tool X. Australian journal of political science, 45(4), pp. 579-604.

Heitshusen Valerie, Young Garry & Wood David M. (2005). Electoral Context and MP Constituency Focus in Australia, Canada, Ireland, New Zealand, and the United Kingdom. American Journal of Political Science, 49(1), pp. 32–45. https://doi.org/10.1111/j.0092-5853.2005.00108.x

Hunt Charles R. & Fontana David (2024). Is Some Politics Still Local? Voter Preferences for Local Candidates. Political Behavior, pp. 1-23.

Huwyler Oliver, Giger Nathalie, Bailer Stefanie, Turner‐Zwinkels Tomas, & Heller Silvan. (2025). Constituency references in social media: MPs’ usage and voters’ reaction. European Journal of Political Research. 64: 1974–1998 doi: 10.1111/1475-6765.70018

Itzkovitch-Malka Reut (2021). MPs’ constituency orientation in the absence of electoral districts: The case of extreme proportional electoral systems. Parliamentary Affairs, 74(2), pp. 421–442.

Jacoby, William, G. (2000). Issue framing and public opinion on government spending. American Journal of Political Science, pp. 750-767.

Jacobsen Roy A. & Larsson Anders O. (2024). Business as Usual? Assessing Amplified Political Posts Across Social Media Platforms During the COVID-19 Pandemic. Social Media+ Society, 10(4).

James Oliver & John Peter (2006). Public Management at the Ballot Box: Performance Information and Electoral Support for Incumbent English Local Governments, Journal of Public Administration Research and Theory, 17(4), pp. 567–580. https://doi.org/10.1093/jopart/mul020 .

Jenkins Henry, Ford Sam & Green Joshua (2013). Spreadable media: Creating value and meaning in a networked culture. In Spreadable media. New York University Press.

Jérôme Bruno & Lewis-Beck Michael S. (1999). Is local politics local? French evidence. European Journal of Political Research, 35(2), pp. 181-197.

Kimberlee Richard H. (2002). Why Don’t British Young People Vote at General Elections? Journal of Youth Studies, 5(1), pp. 85–98. https://doi.org/10.1080/13676260120111788.

Kubin Emily & Von Sikorski Christian (2021). The role of (social) media in political polarisation: a systematic review. Annals of the International Communication Association, 45(3), pp. 188-206.

Lee Benjamin (2014). Window Dressing 2.0: Constituency-Level Web Campaigns in the 2010 UK General Election. Politics, 34, 45 - 57. https://doi.org/10.1111/1467-9256.12029.

Lilleker Darren G., Jackson Nigel & Koc-Michalska Karolina (2015). Social media in the UK election campaigns 2008-2014: Experimentation, innovation, and convergence. In The Routledge companion to social media and politics. Routledge, pp. 325-337.

Marozzo Fabrizio & Bessi Alessandro (2017). Analyzing polarisation of social media users and news sites during political campaigns. Social Network Analysis and Mining, 8(1), pp. 1–13. https://doi.org/10.1007/s13278-017-0479-5

Marres Noortje (2018). Why we can’t have our facts back. Engaging Science, Technology, and Society4, pp. 423-443.

Marres Noortje (2021). ‘No Issues without Media: The Changing Politics of Public Controversy in Digital Societies’, in MEDIA. Intellect, pp. 228–243. Available at: https://intellectdiscover.com/content/books/9781789383263.book-part-023 .

Mason Lilliana (2018). Uncivil Agreement: How Politics Became Our Identity. University of Chicago Press.

Milli Smitha, Carroll Micah, Wang Yike, Pandey Sashrika, Zhaob Sebastian and Dragan Anca D. (2025). Engagement, user satisfaction, and the amplification of divisive content on social media. PNAS nexus, 4(3), pgaf062.

Milazzo Caitlin & Townsley Joshua (2020). Conceived in Harlesden: Candidate-centred campaigning in British general elections. Parliamentary Affairs, 73(1), pp. 127-146.

Møller Hartley Jannie, Bengtsson Mette, Schjøtt Hansen Anna & Sivertsen Morten F. (2023). Researching publics in datafied societies: Insights from four approaches to the concept of ‘publics’ and a (hybrid) research agenda. New Media and Society, 25(7), pp. 1668-1686. https://doi.org/10.1177/14614448211021045

Norri Pippa & Lovenduski Joni (1995). Political Recruitment: Gender, race, and class in the British Parliament. Cambridge University Press.

Ouest-France (2024). Stellantis supprime 200 emplois intérimaires sur le site de La Janais au sud de Rennes. [online] Available at: https://www.ouest-france.fr/economie/entreprises/psa/stellantis-supprime-200-emplois-interimaires-sur-le-site-de-la-janais-au-sud-de-rennes-42905c7e-afc1-11ee-82ae-73e88db45bf9

Parodi Jean-Luc (1978). Les Règles du Scrutin Majoritaire. Projet, 122, pp. 191–200.

Pattie Charles J., Johnston Ronald J & Fieldhouse Edward A. (1995). Winning the Local Vote: The Effectiveness of Constituency Campaign Spending in Great Britain, 1983–1992. American Political Science Review 89(4), pp. 969–983. https://doi.org/10.2307/2082521.

Pattie Charles J., Johnston Ronald (2003). Local battles in a national landslide: constituency campaigning at the 2001 British General Election. Political Geography, 22(4), pp. 381–414. https://doi.org/10.1016/S0962-6298(02)00097-5.

Peinado García Alejandro, Harguindéguy Jean-Baptiste, Cole Alistair, Jiménez Pérez Francisco José & Martínez Cousinou Gloria (2025). Regional ambassadors or state agents? Representing Welsh interests in the British Parliament and Cabinet. Territory, Politics, Governance, 13:5, pp. 597-617, DOI: 10.1080/21622671.2024.2377634

Peterson Erik, & Simonovits Gabor (2018). The electoral consequences of issue frames. The Journal of Politics 80.4: 1283-1296.

Pons Vincent (2018). Will a Five-Minute Discussion Change Your Mind? A Countrywide Experiment on Voter Choice in France. American Economic Review, 108(6), pp. 1322–1363. https://doi.org/10.1257/aer.20160524.

Roemmele Andrea & Gibson Rachel (2020). Scientific and subversive: The two faces of the fourth era of political campaigning. New Media and Society22(4), pp.595-610.

Russell Annelise (2022). Constituent connections: senators’ reputation building in the age of social media. Journal of Information Technology and Politics, 19(2), pp. 180–196. https://doi.org/10.1080/19331681.2021.1949658.

Schrijver Frans J. (2004). Electoral performance of regionalist parties and perspectives on regional identity in France. Regional and Federal Studies 14.2, pp. 187-210.

Schulte-Cloos Julia & Bauer Paul C., (2023). Local candidates, place-based identities, and electoral success. Political Behavior, 45(2), pp.679-698.

Schürmann Lennart (2023). Do competitive districts get more political attention? Strategic use of geographic representation during campaign and non-campaign periods. Electoral Studies, 81, 102575.

Strömbäck Jesper & Kiousis Spiro (2014). Strategic political communication in election campaigns. Political communication 1.8, pp. 109-115.

Taras, David & Davis Richard (2022). Electoral campaigns, media, and the new world of digital politics. University of Michigan Press.

Tavits Margit (2010). Effect of Local Ties on Electoral Success and Parliamentary Behaviour: The Case of Estonia. Party Politics, 16(2), pp. 215–235. https://doi.org/10.1177/1354068809341053

Treré Emiliano (2018). From digital activism to algorithmic resistance. In The Routledge companion to media and activism (pp. 367-375). Routledge.

Tromble Rebekah (2018). Thanks for (actually) responding! How citizen demand shapes politicians’ interactive practices on X. New Media and Society, 20(2), pp. 676–697. https://doi.org/10.1177/1461444816669158.

Viganò Edoardo Alberto (2024). Electoral incentives and geographical representation: Evidence from an Italian electoral reform. Legislative Studies Quarterly, 49(2), pp. 257–287.

Vivyan Nick & Wagner Markus (2016). House or home? Constituent preferences over legislator effort allocation. European Journal of Political Research, 55(1), 81–99.

Vliet Livia, van Törnberg Petter and Uitermar Justus (2020). The Twitter parliamentarian database: Analyzing Twitter politics across 26 countries. PloS One, 15(9), Article e0237073.

Weismueller Jason, Harrigan Paul, Coussement Kristof & Tessitore Tina (2022). What makes people share political content on social media? the role of emotion, authority and ideology. Computers in Human Behavior, 129, 107150. https://doi.org/10.1016/j.chb.2021.107150

Wilson, George (2023). A comparative study of regional-language immersion education in Brittany and Wales. Current Issues in Language Planning, 24(4), pp.418-439.

Silva Bruno Castanho, Schürmann Lennart & Proksch Sven-Oliver (2024). Modulation of democracy: Partisan communication during and after election campaigns. British Journal of Political Science, 54(2), pp. 339–354.

Swalve Tilko, Nyhuis, Dominic, Hönnige, Christoph, Huber Merle & Köker Philipp. (2025). How Centralised is Party Communication on Social Media? Social Science Computer Review, 08944393251399840.

Yoon Nara, Hemsley Jeff, Smith Alexander, Simpson Ellen & Eakins James (2022). Super‐amplifiers! The role of Twitter extended party networks in political elections. Policy & Internet, 14(3), 711-730.

Zittel Thomas (2009). Lost in technology? Political parties and the online campaigns of constituency candidates in Germany’s mixed member electoral system. Journal of Information Technology and Politics, 6(3-4), pp. 298–311.

Zittel Thomas & Gschwend Thomas (2008). Individualised Constituency Campaigns in Mixed-Member Electoral Systems: Candidates in the 2005 German Elections. West European Politics, 31(5), pp. 978–1003. https://doi.org/10.1080/01402380802234656.

Zittel Thomas, Nyhuis Dominic & Baumann Markus (2019). Geographic representation in party-dominated legislatures: A quantitative text analysis of parliamentary questions in the German Bundestag. Legislative Studies Quarterly, 44(4), pp. 681–711. https://onlinelibrary.wiley.com/doi/abs/10.1111/lsq.12238

Haut de page

Notes

1 Formal incorporation into centralised nation-states in the 16th century Brittany in 1532; Wales via the Laws in Wales Acts 1536–1543.

2 Initiated with the 1998 Government of Wales Act and significantly expanded in 2006 and 2017.

3 Conservative and Unionist Party (Conservative), Green Party (Green), Heritage Party (Radical Right), Labour Party (Socialist), Liberal Democrats (Liberal), Plaid Cymru (Regionalist), Propel (Regionalist), Reform UK (Radical Right), Trade Unionist and Socialist Coalition (Radical Left). Source: The Chapel Hill expert surveys.

4 Ensemble - Horizons (Liberal), Ensemble Modem (Liberal), Ensemble - Renaissance (Liberal), Les Republicains (Conservative), Lutte ouvriere (Radical Left), Nouveau Front Populaire - Europe Ecologie Les Verts (Green), Nouveau Front Populaire - La France Insoumise (Radical Left), Nouveau Front Populaire - Parti Socialiste (Socialist), Nouvelle Energie (Conservative), Parti Breton (Regionalist), Rassemblement National (Radical Right). Source: The Chapel Hill expert surveys.

5 Cole (2013) highlights that the concept of the ‘region’ is contested and politically charged, in both Wales and Brittany. In light of this, we avoid using the category of ‘region’ in our labels. Instead, we adopt a more operational central vs. local distinction, where ‘local’ refers to the immediate territorial attachments and language of the constituency, as well as the broader sub-national context, while ‘central’ captures language, topics, and framings associated with central government politics in the UK and France.

6 Figure 4. URLs: https://x.com/Kiera__Marshall/status/1808603753618042886; https://twitter.com/claudia_rouaux/status/1805645142939869502;

7 Figure 5. URLs: https://twitter.com/Kiera__Marshall/status/1804246931939688806

8 Figure 7. URLs: https://x.com/Kiera__Marshall/status/1802426029467672626 and 7 https://x.com/Mathieu_Guihard/status/1804233660616323547

9 Figure 9. URLs: https://x.com/WelshLabour/status/1803498960486891844 https://x.com/Plaid_Cymru/status/1811092789842608305

10 Figure 10. URLs: https://x.com/AcademieDuGallo/status/1808872724325953563; https://twitter.com/Mathieu_Guihard/status/1808911261255508270

11 Figure 11. URLs: https://x.com/WelshLabour/status/1802427582702641316 ; https://x.com/jevershed01/status/1802426256413036666

12 Figure 12. URLs: https://x.com/Plaid_Cymru/status/1808870561566593405

13 Figure 11. URLs: https://x.com/cgtinterim/status/1808470119565115484

https://x.com/HamonValrie1/status/1810342742062424403

14 Figure 14. URLs: https://x.com/rodneyberman/status/1803687539984769081; https://x.com/neiljmcevoy/status/1800448891416199426

15 Figure 15. URLs: https://x.com/hubert_ludo/status/1809147147515785430 and https://twitter.com/AgirPourRennes/status/1805998970549744117

16 Figure 18. URLs: https://x.com/RNBretagne/status/1810348109261885872

17 Figure 18. URLs: https://x.com/Griffpembs13/status/1809653875399442432

Haut de page

Table des illustrations

Titre Figure 1: Social media platform usage by political party in Cardiff and Rennes (across all constituencies).
URL http://journals.openedition.org/reset/docannexe/image/6406/img-1.png
Fichier image/png, 142k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-2.png
Fichier image/png, 81k
Titre Figure 3: Local vs central framing in tweets by political party (Cardiff and Rennes)
URL http://journals.openedition.org/reset/docannexe/image/6406/img-3.png
Fichier image/png, 150k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-4.png
Fichier image/png, 938k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-5.jpg
Fichier image/jpeg, 159k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-6.png
Fichier image/png, 139k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-7.png
Fichier image/png, 141k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-8.png
Fichier image/png, 454k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-9.png
Fichier image/png, 82k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-10.png
Fichier image/png, 733k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-11.png
Fichier image/png, 662k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-12.png
Fichier image/png, 606k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-13.png
Fichier image/png, 62k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-14.png
Fichier image/png, 1011k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-15.png
Fichier image/png, 736k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-16.png
Fichier image/png, 984k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-17.png
Fichier image/png, 52k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-18.png
Fichier image/png, 55k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-19.png
Fichier image/png, 73k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-20.png
Fichier image/png, 172k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-21.png
Fichier image/png, 178k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-22.png
Fichier image/png, 394k
URL http://journals.openedition.org/reset/docannexe/image/6406/img-23.png
Fichier image/png, 419k
Haut de page

Pour citer cet article

Référence électronique

Chiara Poletti et Marco Michieli, « The local in the Digital: comparative insights from Cardiff and Rennes in the 2024 Elections  »RESET [En ligne], 16 | 2025, mis en ligne le 30 janvier 2024, consulté le 17 mai 2026. URL : http://journals.openedition.org/reset/6406 ; DOI : https://doi.org/10.4000/15num

Haut de page

Auteurs

Chiara Poletti

Swansea University

Marco Michieli

Swansea University

Haut de page

Droits d’auteur

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC BY-NC-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search