Advancing the operationalization of national policy styles

While national policy styles have (re)gained academic attention in recent comparative public work, the concept still needs a widely accepted operationalization that can allow collection and analysis of data across contexts while steering away from construct validity threats. We build on Tosun Howlett's (2022) work employ mixed-methods approach, which relies exploratory factor hierarchical cluster analysis. put forth an operationalization, using Bertelsmann's Sustainable Governance Indicators (SGI) as proxies, achieves conceptual clarity distinctiveness, informational robustness, statistical power. Ultimately, we two composite indicators—mode problem-solving inclusiveness—calculate them 41 countries present style classifications based their combinations. report distribution four (administrative, managerial, accommodative, adversarial) conclude with clusters, assessments comparison other classification schemes. 尽管国家政策风格在近期的比较公共政策研究中（重新）获得了学术关注，但此概念仍需要一种被广泛认可的操作化，后者允许跨情境的数据收集和分析，同时避免建构效度(construct validity)威胁。基于Tosun和Howlett(2022)的研究，我们采用包含探索性因素分析和层次聚类分析的混合方法。通过使用贝塔斯曼的可持续治理指标 (SGI)，我们提出一种操作化，后者能实现概念清晰性和独特性、信息稳健性、以及统计能力。最后，我们建构了两个综合指标（问题解决模式和包容性），在41个国家计算这两个指标，并根据指标组合来提出政策风格分类。我们报告了四种政策风格（行政式、管理式、宽松式、对立式）的国家分布，并以聚类分析、稳健性评估、以及与其他国家政策风格分类方案的比较作为结论。 Si bien los estilos de políticas nacionales han (re)ganado la atención académica en el trabajo reciente públicas comparativas, concepto aún necesita una operacionalización ampliamente aceptada que pueda permitir recopilación y análisis datos todos contextos mientras se aleja las amenazas validez constructo. Nos basamos Howlett empleamos un enfoque métodos mixtos, basa factorial exploratorio conglomerados jerárquicos. Presentamos puesta práctica, utilizando Indicadores Gobernanza Sostenible Bertelsmann como sustitutos, logra claridad distinción, solidez información poder estadístico. En última instancia, construimos dos indicadores compuestos: modo resolución problemas e inclusión, calculamos países presentamos clasificaciones basadas sus combinaciones. Informamos distribución cuatro política (administrativo, gerencial, acomodaticio, contradictorio) concluimos con grupos, evaluaciones comparación otros esquemas clasificación estilo nacional. has (Howlett & Tosun, 2019; Leiren et al., 2021; Oztig, 2022; Zahariadis 2021, 2022), analytical progress been hampered by lack systematic applications large number countries. This may be partly due to inability quantitatively operationalize ways permit different contexts. aim change utilizing indicators taken project. They argue convincingly some found ‘good governance’ grouping database, specifically strategic planning consultation, are useful proxies for measuring dimensions styles: mode state-society relations. argument is theoretically empirically beneficial, especially terms quantitative literature forward, recommendation regarding problematic treated fall short operationalizing instead propose new set variables measure styles. not first SGI this purpose, but instead, advance Although retain core value SGI's good governance measures proxy (referred simply hereinafter), chart direction overcomes limitations posed threats, carefully calibrating effects three criteria: After reviewing conceptualizations limitations, examine how our measurement affects classificatory expectations Applying (EFA) (HCA) data, amend demonstrate more should fruitfully used tools long-term patterns policy-making, including analyses. The focuses institutional arrangements govern deliberations actors interact each within broader ideational framework. These interactions provide blueprint through policy-making filtered path dependencies (Pierson, 2004). Hence, modes deliberation—policy styles—affect alongside previous decisions. Tackling Richardson (1982) suggested these interactions: (a) way governments approach problem-solving, either anticipating or reacting problems, (b) government society interact, consensus-building top-down Out two-by-two matrix conceptualization, emerge, under could classified. main issue, however, was Richardson's original framework capture complexities political systems, when it came role bureaucrats experts vis-à-vis public, well degrees inclusion such decision-making. At same time, restricted European or, broadly speaking, western liberal democratic systems. Tosun's (2019) elaboration precisely criticisms, accepting premise proposing looks at key variable expressed high/low inclusiveness societal policy-making. Doing so enables authors apply beyond Europe major developing developed In study interaction trust, al. revisit use reconceptualization springboard expand refine capacity solve problems. focus “ways doing things” idea (Mazey Richardson, 1993) manifests governments' capacity. For instance, if anticipatory reasonable expect sort monitoring implementation words, actual focus, must also circumscribed. administrative interest (high/low capacity) relations inclusiveness). Policy-making strongly affected existing (more than bureaucratic structures, remain largely unchanged) determines government's shaping opportunities available alternatives. But remains qualitative measurement; do any scores style. step forward. However, exclusively does construct's richness. state apparatus include strong capabilities, productive process because implement decisions via apparatus. Gustaffson, (1982, p. 2) categorically assert “also encompasses execution.” State-society relations, per (2019), refer accommodative—or not—patterns mediation. social consultation. What matters goes back (1982). mediation degree institutionally formalized is. On one hand, cases where meaningfully included process, consensus—or, least, mutual understanding—is built formulation adoption stages, aiming reducing potential conflict stage. systems minimal consultation takes place, usually directed top down. addition level there exist example, only institutionalized formal hearings remittances system's transparency, quality information citizens receive participation (e.g., Balla, 2015). Nuances system dictated traditions, historical evolution affect Therefore, inclusiveness, captures broad range citizen group processes, optimal closest conceptualization. Policy necessitates broader, complete feasible data. We, therefore, (2022), suggesting convey robustness strengthen its Taking question best numerically overcome Our rests indicators—we concentrate those, do—contain valuable pertinent solid theoretical basis incorporate, purposes, important numerical “capacity steer [policy] processes success,” they limited temporal geographical coverage problems constructing aggregate individual attributes among others (Schraad- Tischler Seelkopf quoted Croissant Pelke, 2022, 144). realizing indicators, acknowledge imperfect were collected purpose. To answer suited style, need address issues: methodological afflict variables. Construct threat plague every research design: indicator says (Cronbach Meehl, 1955)? GDP power potentially suffers threats entails much economic output. Similarly, highest level. A “way involves execution plans (Richardson, 1982). Composite additionally vulnerable term mathematical combination component (Nardo 2005, 7). Such “send misleading non-robust messages poorly constructed misinterpreted” 6). reason, essential ensures transparency. follow proposed Freudenberg (2003), consists five stages: theory, selection variables, standardization, weighting, robustness. issues, collapse above stages into phases/criteria: distinctiveness clarity, addresses most critical stage, theory. As (2003, 7) asserts: “A needed combine meaningful components weights.” Informational encapsulates all, summary abstract constructs directly measurable. Selecting variable. Statistical incorporates remaining stages. Standardization refers normalizing measured units scales, problem here. weighting conducting tests ensure reliability. departure point enriched refinements (2021, 2022). task inclusiveness. Mode things.” It summarizes specific frameworks. useful, bit nature prerequisites (competencies capabilities) (Wu 2015, 165). competency (legal, administrative, human) place coordinated (at right time) perform functions effectively (Peters, go conceptualization (planning adoption), dimensions. Unlike offer numerical, dichotomous, nuanced explicitly construct. think plan courses action; ministerial politics adopting particular solutions problems; course execute them. Conversely, indicates seat table formalization (Zahariadis conceptually distinct style: whereas solving feature state's apparatus, foregrounds ability formally incorporate input decision-making processes. cross-nationally, turn SGI, did, select issues. project normative notion make policies challenges face—globalization adversities, cleavages, resource demographic shifts—while considering (Bertelsmann Stiftung, aims inform OECD cross-national comparisons customized forming pillars (or indices): performance, robust democracy, (Schraad-Tischer Seelkopf, study, pillar, guided question, “[h]ow capacities OECD/EU between actors?” (Schiller 7), concept. conceptualizes both act institutions influence what does. pillar divided dimensions: executive accountability, various exist; additional More specifically, capacity, interministerial coordination, implementation. correspond planning, adoption, delivery services. communication, citizens' participatory competence. except are, se, indicators. Here break (2022); entire rather selecting components. recognize drawback occasional aggregation higher-level concepts without adequate justification (Croissant 155). Yet, reasonably designed many facets possible. cover part elements process: Strategic components: expert advice. concerns forms play view challenges. hand-in-hand second component, advice, regularly consider advice Interministerial coordination negotiating line ministry proposals against priorities set. expertise offices; involvement ministries; cabinet committees filter out settle issues before meetings; extent officials civil servants coordinate drafting ministries reach coordination; existence mechanisms, digitalization coordination. turn, links execution. Implementation effective policies. Its achieve objectives; mechanisms compliance; instruments monitor agencies; tasks allocated subnational adequately funded; decentralization government; realize services, finally, deal powerful lobby groups comes groups' process. Using allows us nuance problem-solving. overlaps substantially 2022) closer conventional parse levels attendant bureaucracies, relationships them—both informal—thus boosting increases validity. exhaustive aspects perspectives system. interpenetration Societal assesses successful inviting opinions organizations trade unions, business associations, religious communities, others. communication whether communicates goals factually coherent way. assess informed publishes strengthens hold accountable. Even though citizenry informed, increase sense, constructed, statistically valid reliable? variable, linear Inclusiveness competence, communication. sufficiently assign weights selected variable's mean Table 1 provides descriptive statistics correlations Beyond usual centrality dispersion, shape (normality) (Table 1). Skewness kurtosis because, certain values, violate assumption frequently statistics, t-tests ordinary least squares regression, leading high incidence Type I errors (Cain 2017). Shapiro–Wilk test normality yields W statistic 0.968, significant. addition, joint skewness nonsignificant χ2 coefficients 1.40 1.49, respectively. normally distributed. look differences time last point: Do values significantly over medium term? If not, theory predicts visually verify, gains traction. supplement visual inspection, assessment necessary. all six (three inclusiveness) divide sample year groups: 2014, 2017, 2020. know glacial countries, significant short- medium-term sample—3 6 years, run MANOVA equality means groups, assuming homogeneity. Statistically insignificant indicate no substantial variance groups. then heterogeneity Again, Wald's likelihood-ratio results inspection generalized medium-term, find, short-term. There average, temporally stable change. now creation. First, correlate related convergent Second, conduct EFA ascertain loadings onto latent factors. load factor, represents gained credibility unidimensionality groupings. Finally, calculate scale reliability coefficient, Cronbach's α, internal consistency measurement. Researchers typically establish correlating associated (convergent validity) vary independently (discriminant (Campbell Fiske, 1959). Unfortunately, cannot possibly discriminant several explore related. Two organizational reform evidence-based instruments. note items identical gain further confidence having obstacles. positively (r = 0.717). Countries higher political, tend willing structure governance. Their appear able performance appropriate. relates equally 0.729). Governments consult seek encourage greater likely regulatory impact exercises (RIA). logical relationship exists RIAs involve and/or collaboration stakeholders transparent “the providing empirical about reaction individuals regard change” (SGI, Public seem hand-in-hand. dimensionality. sense preconceived notions factors will (Gorsuch, phases: suitability, extraction, retention, rotation, interpretation. suitability Watkins (2018, 226) puts it, “it […] verify intercorrelated justify analysis.” ways. correlation brief shows sufficient intercorrelation 0.50. Bartlett's sphericity tells generated identity matrix, is, off-diagonal randomly zero. coefficient significant,