Positioning analysis of Spanish politicians through their Twitter posts versus Spanish public opinion

The study addresses how political communication on social networks relates to public opinion on public and fiscal policies in Spain. In a context of growing digitalization and the mediatisation of politics, Twitter has become a central medium for political messaging. The research seeks to fill gaps in the literature by jointly analyzing (1) the positioning in politicians’ Twitter discourse and (2) whether such positioning is mirrored in citizens’ evaluations of public services and fiscal policies according to political affinity. The work covers the period 1 June–31 July 2021 and considers a broad spectrum of parties and leaders (including special treatment for ERC due to leadership circumstances). The research questions are: RQ1: How do politicians position themselves through their social media posts? RQ2: Are there significant differences in citizens' opinions on satisfaction with public services depending on political affinity? RQ3: Are there significant differences in citizens' opinions on tax policy depending on political affinity? The dual prism enables a parallel analysis of online leaders’ discourse and offline citizens’ opinions to assess potential alignment or divergence in positioning.

The theoretical framework situates the study within ICT-driven changes to communication and the role of social media in political campaigns. Prior research shows parties and leaders use social media strategically, with Twitter particularly salient in Spain. Social networks can amplify ideology, create echo chambers, and contribute to polarisation and self-esteem among users. Polarisation is discussed as both a social and discursive phenomenon with implications for democratic dialogue, with Spain exhibiting high affective and partisan/territorial polarisation in recent studies. The literature also notes voter tendencies to seek ideologically congruent information online. Regarding public services, prior work suggests satisfaction varies by ideology, and Spaniards often prefer better management over increased taxes, with right-leaning citizens more skeptical of higher tax burdens. Differentiation around fiscal policies, notably redistribution and defence/security, has increased, particularly during COVID-19. From this framework, the authors propose hypotheses: H1a: Politicians’ Twitter discourse positioning differs significantly; H1b: Positioning is polarized between right/left and between constitutionalist/pro-independence. H2a: Citizens’ opinions on public services function show significantly different positioning by party; H2b: These opinions are polarized between right/left and between constitutionalist/pro-independence. H3a: Citizens’ opinions on the allocation of resources to public services show significantly different positioning by party; H3b: These opinions are polarized between right/left and between constitutionalist/pro-independence.

Design: Exploratory review followed by correlational analyses with a dual qualitative–quantitative approach. Scope: Top 10 Spanish political parties by seats (2019 general elections). Period for both components: 1 June–31 July 2021. Qualitative (leaders’ Twitter discourse): - Data: 1,553 tweets posted by 11 political leaders (covering 10 parties; ERC includes both Oriol Junqueras and Gabriel Rufián due to the former’s imprisonment until 23 June 2021; Podemos leadership change to Ione Belarra noted). - Collection: Twitter API scraping of leaders’ personal accounts. - Analysis: Semantic analysis followed by Multiple Correspondence Factor Analysis (MCFA) to create positioning maps using Le Sphinx software. Quantitative (citizens’ opinions): - Data source: CIS “Public Opinion and Fiscal Policy Survey” July 2021. - Sample: N=2,849 Spanish citizens, 18+; simple random sampling via landline (28.3%) and mobile (71.7%); quotas by sex, age, and proportional affiliation; fieldwork 21–29 July 2021 via CATI. - Software: SPSS v27 and DYANE v3. - Analyses: Correspondence Factor Analysis (CFA) linking political affinity (Q32: intended vote if elections were held tomorrow) with (1) satisfaction with functioning of nine public services (Q06; 1=very satisfactory to 5=not at all satisfactory), and (2) perceptions of resource dedication across 15 services/benefits (Q08; 1=too many, 2=necessary, 3=too few). - Variables: Public services for satisfaction: Education, Healthcare, Pension Management, Justice Administration, Public Safety, Social Services, Public Transport, Public Works, Aid to Dependent Persons. Resource dedication: Education, Public Works, Unemployment Protection, Defence, Public Safety, Health, Culture, Housing, Justice, Social Security/Pensions, Transport and Communications, Environmental Protection, Development Cooperation, Research in Science and Technology (R&D), Aid to Dependent Persons. - Reported model metrics include eigenvalues and contributions to inertia for extracted axes in each analysis.

Qualitative (leaders’ Twitter positioning): - MCFA indicates a highly significant relationship between leaders and most-mentioned words (P<0.01; Chi-square=4208.0; df=1000). - Inertia explained: Axis 1 ≈ 29%, Axis 2 ≈ 17%; first five axes cumulatively ≈ 78% (Table 2 also reports contributions: F1 28.57%, F2 16.71%, F3 12.29%, F4 10.68%, F5 9.63%). - Axis 1: clear left–right polarisation. Left side includes Pedro Sánchez (PSOE), Ione Belarra (UP), Iñigo Errejón (Más País) and words like country, woman, social, equality, recovery, new, society. Right side includes Pablo Casado (PP), Inés Arrimadas (Cs), Santiago Abascal (Vox) and words like government, court, law, plotter, Spaniard, national. - Axis 2: clear constitutionalist–pro-independence polarisation. Upper side features pro-independence leaders (Oriol Junqueras, Gabriel Rufián, Andoni Ortuzar, Arnaldo Otegi, Carles Puigdemont) and words such as independence, right, freedom, able; lower side includes Sánchez (with today, agreement, love) and on the right lower side Casado/Arrimadas with unity, Spanish, national, and Spain/family centered lower. - Hypotheses: H1a and H1b accepted (leaders are clearly and polarisedly positioned consistent with ideologies). Quantitative (voter positioning—satisfaction with public services, CFA1): - Scale: 1=very satisfied, 5=not at all satisfied (higher values=dissatisfaction). - Inertia explained: Axis 1=46.39%, Axis 2=34.29% (total ~80.68%). - Axis 1 associates higher dissatisfaction with Justice Administration; opposite side associates with Health and Social Services satisfaction. PNV voters align with greater dissatisfaction with justice and more satisfaction with health/social services/public transport; JxCat next closest. Más País aligns opposite. - Axis 2 defined by Education (positive) vs Public Transport (negative). PNV, followed by Vox and PP, align on the positive side; JxCat and ERC on the opposite. - Patterns: Strong affinities between Vox and PP (more dissatisfied with education); PSOE and UP more centered. No significant left–right polarisation; clearer differences between constitutionalist vs pro-independence. - Hypotheses: H2a accepted (different positioning across parties), H2b not supported (no left–right polarisation; differences clearer between constitutionalist and pro-independence). Quantitative (voter positioning—resources dedication, CFA2): - Inertia explained: Axis 1=77.38%; Axis 2=12.92% (Total inertia 0.004875; Chi² 1.8353). - Axis 1 is dominated by Defence spending (explains ~70.06% of Axis 1 inertia), followed by Security. - Vox and PP voters are closest to the position that too few resources are devoted to Defence and Security; ERC and EH Bildu are in the opposite position. Axis 2 shows no significant differences. - Hypotheses: H3a accepted (different positioning by party), H3b not supported (no clear left–right polarisation; larger differences constitutionalist vs pro-independence). Overall: Leaders’ online discourse is strongly polarised along left–right and territorial lines; voters’ offline opinions show fewer, more nuanced differences, with notable divides on defence/security spending and constitutionalist vs pro-independence identities rather than a simple left–right split.

The findings address the research questions as follows: - RQ1: Politicians’ Twitter discourse reveals clear ideological and territorial polarisation. Left vs right differentiation is strong, with distinct clusters of frequently used terms aligned with party ideologies; a second dimension separates constitutionalist from pro-independence leaders. - RQ2: Citizens’ evaluations of public service functioning differ by party affinity, validating heterogeneity across electorates. However, these differences are not polarised along the left–right axis. Instead, the most salient divide is constitutionalist vs pro-independence, with additional regional distinctions (e.g., Basque vs Catalan parties) and specific issue-based dissatisfaction (e.g., PNV with justice; ERC/JxCat with transport). - RQ3: On allocation of resources, voters exhibit clear differentiation by party, especially on defence and security (PP/Vox vs pro-independence parties), yet again not a simple left–right polarisation. Implications: The contrast between highly polarised online leader discourse and comparatively moderate, issue-specific citizen opinions suggests that social media may accentuate elite-level polarisation more than mass-level divides. Territorial identities appear particularly influential in structuring public opinion differences in Spain. These insights inform political communication strategies, indicating that while polarised messaging may mobilize bases online, offline voter attitudes are more nuanced and issue-dependent.

The study demonstrates that Spanish political leaders’ Twitter communications are clearly and polarisedly positioned along left–right and constitutionalist–pro-independence dimensions. In contrast, citizens’ offline opinions on public service performance and resource allocation show fewer and more nuanced differences, with significant divergence primarily around defence/security spending and territorial (constitutionalist vs pro-independence) identities rather than a straightforward left–right cleavage. The work contributes by jointly mapping elite online discourse and mass offline attitudes within the same time window, offering guidance for party differentiation and message strategy. Future research should extend analysis longitudinally (panel designs), broaden geographical scope beyond Spain, and incorporate additional variables to capture other relevant dimensions of positioning.

• Cross-sectional design: Leaders’ online discourse and citizens’ offline opinions were analyzed in parallel over a short time window (June–July 2021), limiting causal inference and temporal dynamics; a longitudinal/panel design is suggested. - Geographic scope: Focus on Spain limits generalizability; multi-country studies are recommended. - Measurement scope: A limited set of variables was used; future work could add more items to capture additional facets of positioning and policy attitudes.