Populism and Economic Freedom in European Democracies: Evidence from a Two-Way Fixed Effects Panel Analysis, 2001–2023

Populism and Economic Freedom in European Democracies: Evidence from a Two-Way Fixed Effects Panel Analysis, 2001–2023 | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Populism and Economic Freedom in European Democracies: Evidence from a Two-Way Fixed Effects Panel Analysis, 2001–2023 Constantinos Saravakos This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9451071/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This paper seeks to examine the relationship between populism and economic freedom across 31 European democracies, during 2001–2023, extending the research by Celico and Rode ( 2023 ) and complementing Bergh and Kärnä ( 2024 ). The paper contributes three novel dimensions: simultaneous ideological decomposition (radical right vs. radical left); sub-component regressions for all five EFW areas; and a backsliding heterogeneity split (Hungary from 2010, Poland from 2015). Aggregate populism vote share shows no significant association with annual EFW changes, however, when decomposed by ideological host, radical right populism exhibits a significant negative direct association (WCB p = 0.030) attenuated by stronger judicial constraints, concentrated in stable democracies and near-zero during backsliding episodes (rr × backsliding interaction p = 0.016). This finding is consistent with the mixed institutional legacy of Orbán- and PiS-style governance. Among sub-components, Freedom to Trade Internationally shows the strongest negative association with radical right populism (WCB p = 0.063), consistent with the protectionist orientation of contemporary European far-right platforms. The null result for radical left populism may reflect a baseline-shift mechanism: mainstream and radical left parties were already converged on anti-trade positions, leaving no measurable policy distance, asymmetrically reinforced external conditionality that disciplined left-populist parties in ways that had no equivalent for right-populist incumbents. A Dumitrescu–Hurlin panel Granger test at 2, 3, and 5 lags consistently fails to reject the null of no Granger causality from populist vote shares to EFW changes, an indication of strong correlational associations, but not a causal effects identification. populism economic freedom radical right panel data democratic backsliding Figures Figure 1 Figure 2 1. Introduction The past two decades have witnessed a sustained rise of populist parties across European democracies, transforming them from a marginal political phenomenon into a defining feature of electoral competition. From Hungary’s Fidesz to Italy’s Lega, Poland’s Law and Justice (PiS), and Sweden’s Democrats, far-right formations have managed to expand their vote shares substantially. At the same time, left-populist formations—from Syriza in Greece to Podemos in Spain—have gained significant electoral support, particularly following the 2008–2010 financial crisis, but for a shorter period of time. By 2023, the mean populism vote share across the 31 European democracies in the sample had reached approximately 25 per cent, nearly double the level observed in 2001. This growth is primarily driven by far-right parties, whose aggregate vote shares have grown from roughly 7 per cent in 2001 to over 17 per cent by 2023, while radical left shares have remained more stable at approximately 6–8 per cent throughout the period, with a peak during the 2012–2019 time frame. 1 Whether this electoral transformation carries consequences for economic institutions and economic freedom has been a question of both academic and policy relevance (Funke, Schularick and Trebesch 2023 ). Economic freedom—as measured by the Economic Freedom of the World (EFW) index of the Fraser Institute—captures the degree to which individuals and enterprises operate in an conducive environment, which protects property rights, and secure sound monetary policy, limited and non-distorting government, open international trade, and limited regulation. Cross-country evidence consistently associates higher EFW scores with better economic outcomes, including higher income, faster growth, reduced poverty, and stronger investment (Lawson 2022 ; Berggren 1999 ; de Haan and Sturm 2000 ). The question of whether the populist wave threatens these institutional arrangements is therefore of considerable importance. The theoretical priors are ambiguous, as populism, regardless of ideological host, is primarily grounded on a conception of “the will of the people” that is inherently hostile to institutions of counter-majoritarian constraints, such as judicial independence, central bank autonomy, regulatory independence, which are themselves the underpinning of economic freedom (Mudde and Rovira Kaltwasser 2017 ; Levitsky and Ziblatt 2018 ). However, the expected direction of institutional change depends critically on whether populists are in government (and can directly dictate public policy) or merely in opposition with rising vote shares (in which case the primary mechanism is signaling-driven uncertainty and reputational effects on institutions). This paper’s empirical design captures the vote-share channel, which includes both the anticipatory effects of populist electoral strength and the marginal influence that large opposition blocs exercise on incumbents’ policy calculus. The distinction between these channels is elaborated in Section 2.4 and is important for interpreting the magnitude of the estimated associations. The questions is not new in institutional economics, as there are two closely related recent studies bracketing this paper’s contribution. Celico and Rode ( 2023 ) use a global panel of electoral democracies from 1970 to 2019 and find that aggregate populism in government is significantly negatively associated with EFW changes for non-OECD countries, while in OECD settings, with strong institutional capacity, the association is largely mediated by political constraints and government ideology. Bergh and Kärnä ( 2024 ), working in the complementary reverse direction, find that in European democracies weaker rule of law (Area 2 of EFW) is significantly associated with higher right-wing populism vote shares, however changes in aggregate economic freedom do not predict changes in populism vote shares. This paper sits at the intersection of these two studies, as elaborated in Section 2.3 . The results confirm that aggregate populism vote share has no statistically significant association with overall economic freedom changes in European democracies, which are institutionally more developed compared to o global panel dataset, a finding consistent with what Celico and Rode ( 2023 ) suggest for OECD-type settings. When decomposed by ideology, radical right populism vote share exhibits a significant negative direct association (p < 0.05) attenuated by stronger judicial constraints. Among sub-components, Freedom to Trade Internationally shows the strongest negative association with radical right populism—marginally significant at the 10 per cent level under the paper’s preferred wild cluster bootstrap inference (WCB p = 0.063; conventional clustered p = 0.041 with EU membership control; conventional p = 0.070 without). Results are robust to alternative specifications. I organise the rest of the paper in the following way. Section 2 reviews the relevant literature. Section 3 describes the data and variables. Section 4 presents the empirical strategy. Section 5 reports results. Section 6 discusses findings and limitations. Section 7 concludes. 2. Literature Review 2.1 The Concept of Economic Freedom and Its Determinants The EFW index aggregates 42 components across five areas: (1) size of government, (2) legal system and property rights, (3) sound money, (4) freedom to trade internationally, and (5) regulation, while a substantial body of empirical research shows that higher EFW scores are robustly associated with higher per capita income, faster long-run growth, lower inflation, and better institutions (Gwartney, Lawson and Holcombe 1999 ; de Haan and Sturm 2000 ; Hall and Lawson 2014 ; Lawson 2022 ). The determinants of EFW changes over time have received less systematic attention, though studies have identified democratic institutions (Bjørnskov and Rode 2020 ), government ideology (Bjørnskov and Potrafke 2012 , 2013 ), income and development level (Rode and Gwartney 2012 ), and trade openness as relevant correlates. Bjørnskov and Potrafke ( 2012 , 2013 ) establish an important baseline: governments of the political right are systematically associated with higher subsequent EFW scores, while left-wing governments are associated with reductions (i.e. less economic freedom), particularly in government size and regulation components. This ideological baseline is important for this study because the relationship between populism and economic freedom cannot be disentangled from the ideological character of the populist movement in question. 2.2 Populism: Conceptualisation and Measurement The concept of populism has been operationalised through several approaches. Binary classifications (Rooduijn et al., 2019 ; Heinö 2024 ) offer broad coverage but they fall short on information on the extent and the intensity of populist phenomenon. Celico, Rode and Rodríguez Carreño ( 2022 ) use machine-learning random forest methods to extend expert-survey data to 1,920 parties across 163 countries from 1970 to 2019, yielding a continuous 0–10 scale capturing the degree of populist discourse. This study uses Timbro’s Authoritarian Populism Index (TAP, Heinö 2024 ) dataset, which provides continuous vote share data decomposed into radical right and radical left components across European democracies. This operationalisation captures electoral support rather than parliamentary representation, which allows a direct side-by-side comparison of right versus left populism effects without relying on a scalar ideology moderator. Crucially, electoral strength need not translate into government participation to generate institutional consequences. The reason is that a large populist bloc in opposition is able to constrain incumbent policy space, shift the median voter calculus, and create anticipatory uncertainty among investors and trading partners—all mechanisms through which rising vote shares can affect economic freedom indicators well before any formal transfer of executive power. TAP vote share data are broadly consistent with the PopuList coding (Rooduijn et al., 2019 )—both draw on similar expert classifications of party families. 2 A related definitional caveat is the relationship between radical populism and Euroscepticism. A significant share of parties coded as radical right in TAP are simultaneously Eurosceptic (opposing deeper European integration, opposing the single market, or seeking treaty revision). To the extent that Eurosceptic rhetoric shapes trade-policy expectations—over and above the protectionist motivation captured by the Area 4 channel—the radical right vote share variable partially proxies Eurosceptic sentiment, to a great extent. Disentangling the protectionist from the purely sovereignty-based Eurosceptic component would require a finer-grained party-ideology measure than is available in the TAP dataset. Therefore, I flag this as a scope limitation, as the results of Freedom to Trade Internationally area may reflect a combination of protectionist policy signalling and Eurosceptic-driven uncertainty about the future trade architecture of the country. An important caveat concerns within-category heterogeneity. Parties classified as ‘radical right’ span a wide range of economic programmes: Marine Le Pen’s Rassemblement National has historically combined welfare chauvinism with economic interventionism, while Viktor Orbán’s Fidesz has pursued a more orthodox fiscal consolidation alongside selective protectionism and state-directed industrial policy. Similarly, radical left parties extend to a wide range, from moderate social democrats to more genuinely post-capitalist formations. The radical right and radical left vote share variables therefore aggregate over heterogeneous programmes, and the estimated coefficients capture mean effects within each ideological family. This is a limitation that party-level or policy-specific analysis would help address. 2.3 Populism, Institutions, and Economic Freedom The empirical study of populism’s economic consequences is a relatively recent enterprise. Funke, Schularick and Trebesch ( 2023 ) trace 51 populist episodes since 1900 and find that countries under populist governance suffer substantially larger long-run economic costs—in terms of GDP per capita, trade openness, and institutional quality—with costs compounding over successive terms. More specifically about economic freedom, Rode and Revuelta ( 2015 ) provide the first systematic cross-national study, finding that populism in government is negatively associated with EFW across a sample weighted toward left-populist cases in Latin American. Celico and Rode ( 2023 ) revisit this question with a more comprehensive sample and more sophisticated populism measure, and their key finding is that the association between populism and economic freedom (proxied by EFW) is significant and negative for the full democratic sample and for non-OECD countries, but in OECD countries the negative association is largely mediated by political constraints and government ideology. They interpret this as reflecting institutional safeguards, since in countries with stronger checks and balances, populist governments find it harder to erode economic freedom. Related evidence comes from Stöckl and Rode ( 2021 ), who find that negative financial market effects of populist elections are primarily driven by left-wing populism. Bennett et al. ( 2021 , 2023 ) document that populist governments are associated with institutional erosion, corruption, and weaker property rights, with political constraints limiting these effects. This pattern is consistent with the concept of executive aggrandizement—the incremental, legally cloaked erosion of counter-majoritarian institutions by elected executives who retain formal democratic legitimacy while systematically dismantling its substance (Bermeo 2016 )—rather than with outright coups or extra-constitutional seizures of power. An important complementary strand examines the exact reverse direction. Bergh and Kärnä ( 2024 ), in the Handbook of Research on Economic Freedom, address whether economic freedom predicts populism for 33 European democracies over 1980–2020. Their panel fixed-effects analysis reveals one highly robust finding: legal system and property rights (Area 2 of the EFW index, i.e. rule of law) is significantly and negatively correlated with right-wing populism vote shares, while no other EFW area manged to achieve significance. For left-wing populism, no EFW component is significantly associated with vote shares. Examining changes over time, they find no correlation between changes in aggregate economic freedom and populist vote shares. This null result provides partial reassurance—discussed in Section 6.4 —that the reverse causality channel in this specification is not sufficiently operative. The two studies—Celico and Rode ( 2023 ) and Bergh and Kärnä ( 2024 )—frame the contribution of the present paper, as it differs from both along three distinct dimensions. First, on sample scope and direction of inquiry: unlike Celico and Rode’s global panel of 2,367 observations, this paper focuses specifically on European democracies, which serve as a) the institutional setting directly relevant to the contemporary populist wave and b) as the set of countries with the most developed check and balances mechanisms in western democracies—while using a longer time horizon (to 2023) than Bergh and Kärnä (to 2020); and, like Celico and Rode, it models populism as a predictor of EFW changes, complementing rather than replicating Bergh and Kärnä’s reverse-direction framework. Second, on the treatment of ideology: both prior papers use a single scalar measure or separate single-equation estimates for right versus left, whereas this study includes radical right and radical left vote shares simultaneously in the same equation. Third, and most distinctively, I conduct sub-component regressions for each of the five EFW areas as separate outcomes—asking not only whether aggregate economic freedom is affected but which specific institutional channels are identified as the most sensitive to populist electoral pressure. This sub-component analysis is, to the best of author’s knowledge, absent from earlier research and from the broader comparative literature on populism and economic institutions. 2.4 Theoretical Mechanisms: Policy Channel and Signalling Channel A theoretical concern with vote-share specifications requires explicit acknowledgment, since the effect of populist electoral strength on economic freedom may operate through two conceptually distinct mechanisms, which the current design captures jointly rather than separately. The first is the direct policy channel: populists who enter the parliament and cabinets, can affect policy making by legislate policies—protectionist tariffs, central bank subordination, property right abridgements, expanded transfers—that mechanically alter the EFW sub-scores directly. The second is a signalling channel: rising populist vote shares create anticipatory uncertainty among investors, trading partners, and institutional actors even before populists reach office, generating reputational and expectation effects that can affect economic freedom indicators. Such effects could be the widening of sovereign spreads, reduced foreign direct investment, or voluntary regulatory forbearance by incumbents seeking to neutralise the populist threat. The empirical design—using continuous vote shares (not government participation) as the key regressor—is best interpreted as capturing the combined electoral-strength effect, including both the direct policy effects of populists in government and the signalling effects of their presence as a large electoral force. I include a robustness check using a binary Populists in Government indicator (Section 4.4 ), which isolates the direct governance channel. The Discussion interprets results in light of both mechanisms, treating them as one. The public choice literature offers a further theoretical mechanism for the trade freedom channel specifically: protectionism as coalition-building, where radical right parties favour domestically organised producers over diffuse consumer interests—a mechanism that can operate both through direct policy (tariff legislation) and through electoral signalling to incumbent governments facing populist competition (Autor et al., 2020 ; Guriev and Papaioannou 2022 ). 3. Data and Variables 3.1 Dependent Variable: Economic Freedom of the World The primary dependent variable is the annual change in the Economic Freedom of the World (EFW) index (Gwartney, Lawson, Hall and Murphy 2023 ), available annually from 2000 onwards on a 0–10 scale. I model the change ΔEFW = EFW_t − EFW_{t-1} as the dependent variable rather than the level, following Celico and Rode ( 2023 ) and Rode and Revuelta ( 2015 ). For sub-component analysis, I use ΔEFW Areas 1–5 as separate outcomes. Raw controls for sub-component regressions are drawn from the very EFW 2025 Master Index Dataset: government consumption as a share of GDP (raw component of Area 1, entered as its natural logarithm given the right-skewed distribution); the annual CPI inflation rate (raw component of Area 3, entered as log(1 + π/100) × 100 to accommodate the 62 country-year observations with negative inflation values while compressing the high-inflation tail); and the mean applied tariff rate (raw component of Area 4, entered in percentage levels). The tariff rate is not log-transformed because several country-year observations record zero values; a log(1 + tariff) transformation was tested in robustness and does not end up with significant different results. 3.2 Populism Variables Populism data come from the Timbro’s Authoritarian Populism index (TAP, Heinö 2024 ) dataset covering European democracies. I use: (1) overall populism vote share (POP_VOTE), the combined vote share of all populist parties; (2) radical right vote share (RRIGHT_VOTE); and (3) radical left vote share (RLEFT_VOTE). Vote shares are recorded at election years and carried forward until the next election. In the sample, mean overall populism vote share is 19.3 per cent (SD = 15.2 pp, as reported in Table 1 ), rising from approximately 12 per cent in 2001 to 25 per cent by 2023, with the increase primarily driven by radical right electoral support. Radical right parties average 12.1 per cent (SD = 13.4 pp); radical left parties average 6.6 per cent (SD = 9.5 pp). Populists participate in government in approximately 21 per cent of country-year observations. TAP records vote shares for parties that meet the dataset’s populism-classification threshold; for country-years where no qualifying radical right or radical left party was present, I assign a vote share of zero, as the most methodologically appropriate coding when the absence of a qualifying party implies zero populist electoral mass. Under this zero-fill convention the full (N = 713) and sub-component (N = 713) analytic samples are identical: all 31 × 23 country-year cells contain complete populism vote share data. A covariate balance check confirms that the two samples are statistically indistinguishable across all key covariates (see Appendix Table A1). The note that TAP is broadly consistent with the PopuList coding (Rooduijn et al., 2019 ) extends to Eurosceptic parties as well: parties coded as radical right or radical left in TAP typically appear under closely related headings in PopuList’s Eurosceptic category, though the two databases draw different party-family boundaries and a formal overlap matrix has not been computed for this paper. I also need to clarify that the carry-forward convention warrants explicit theoretical justification. What this specification measures is electoral strength as an institutional signal: the 33 per cent vote share of the Rassemblement National in France in 2022 does not cease to ‘press’ on the policy calculus of the Macron government in 2023 and 2024—incumbents adjust continuously in anticipation of a large opposition bloc, not only at election years. This is precisely the signalling channel outlined in Section 2.4 . The limitation of the approach is that it does not capture within-term polling fluctuations, which would constitute a higher-frequency measure of populist pressure. 3 3.3 Institutional Controls I use the V-Dem Judicial Constraints index (Coppedge et al., 2023 ) as the primary institutional moderator. The Judicial Constraints (JC) index measures the extent to which courts constrain executive authority, ranging from 0 to 1. I think this measure fits better because it is theoretically more precise for this paper’s question, which concerns the mechanism through which institutional constraints moderate the relationship between populism and EFW, operates primarily through the judiciary’s capacity to prevent executives from dismantling property rights and monetary independence. In the sample, the mean Judicial Constraints index is 0.903 (SD = 0.082), with considerable within-country variation over time, particularly in countries experiencing democratic backsliding. Critically, the interaction between populism and judicial constraints is primarily identified from countries with meaningful JC variation over the sample period—most prominently Hungary and Poland, where the JC index declined significantly after 2010 (Hungary) and 2015 (Poland) as the respective governments moved against judicial independence. 3.4 Additional Controls I include the lagged EFW level (or sub-area level) and the lagged natural logarithm of GDP per capita in PPP terms (2021 international dollars) as standard controls. In addition, I include the lagged EU membership dummy (1 = EU member in year t − 1, 0 otherwise), which takes value 1 continuously for longstanding members and transitions from 0 to 1 at accession for the 14 countries that joined the EU during the panel period (including Bulgaria and Romania in 2007, Croatia in 2013, and the 2004 cohort). Although the EU dummy is largely absorbed by country fixed effects for longstanding members, it captures accession-related policy transitions for the 14 countries with within-sample variation. For Sound Money (Area 3), I note that EU membership is an imperfect proxy for monetary regime: Eurozone members face ECB-disciplined inflation, whereas non-Eurozone EU members retain independent monetary policy. The EU dummy thus conflates Eurozone and non-Eurozone membership for Area 3 analysis, and a more granular monetary-regime indicator would be advisable in future extended work. The EU dummy is not significant in any main specification. 3.5 Sample The intersection of the TAP dataset with the EFW panel yields a sample of 31 European democracies from 2001 to 2023, producing up to 712 country-year observations after lagging (Models 1–2) or 663 observations (Models 3–4, where the right/left decomposition is available). Twenty-six of the 31 countries are OECD members. This homogeneous sample of institutionally mature high-income European democracies allows to focus on a demanding test for the populism–economic freedom nexus: if effects are present here, they persist despite strong institutional safeguards. Importantly, the homogeneity of the sample also implies attenuation bias: coefficients estimated on a sample with limited variance in both the JC index (confined to a narrow high range) and the EFW outcome (compressed by institutional stability) are likely downward-biased in absolute value relative to what would be found in a more diverse global panel. Namely, this test can be treated as a conservative estimation for our results. Descriptive statistics are in Table 1 . Table 1 Descriptive Statistics Variable N Mean SD Min Max Δ EFW Overall (dep. var.) 713 0.005 0.138 −0.835 0.850 Δ Area 1: Government Size 713 −0.005 0.246 −1.320 0.977 Δ Area 2: Legal System & Property Rights 713 0.007 0.111 −0.353 0.855 Δ Area 3: Sound Money 713 0.020 0.447 −2.456 3.263 Δ Area 4: Freedom to Trade Internationally 713 0.005 0.228 −1.236 1.471 Δ Area 5: Regulation 713 0.000 0.211 −1.697 0.717 EFW Overall (level) 713 7.623 0.424 5.753 8.670 Populism Vote Share (%, t − 1) 712 19.281 15.211 0.000 69.500 Radical Right Vote Share (%, t − 1) 663 12.132 13.396 0.000 69.400 Radical Left Vote Share (%, t − 1) 667 6.598 9.483 0.000 45.100 Populists in Government (0/1) 713 0.208 0.406 0.000 1.000 EU Membership (0/1, t − 1) 712 0.776 0.417 0.000 1.000 Judicial Constraints Index (V-Dem, t − 1) 713 0.903 0.082 0.507 0.988 Liberal Component Index (V-Dem, t − 1) 713 0.905 0.068 0.658 0.984 Log GDP per capita PPP (t − 1) 713 10.701 0.427 9.439 11.840 Log Govt Consumption (% GDP, t − 1) 712 3.269 0.219 2.084 3.652 CPI Inflation Rate (%, t − 1) 712 2.889 3.284 −4.480 34.468 Note: Sample covers 31 European democracies, 2001–2023. Δ denotes annual first difference. All lagged variables from t − 1. Sources: Fraser Institute EFW 2025; TAP; V-Dem v16; World Bank WDI. 4. Empirical Strategy 4.1 Identification Approach The empirical design exploits within-country, within-year variation after removing country-specific time-invariant characteristics and common time trends. Concretely, the two-way fixed effects (TWFE) estimator demeans each variable by subtracting the country mean, the year mean, and adding back the overall mean (the within transformation). The remaining variation—‘within-country variation relative to the cross-country and cross-year patterns’—is what identifies the coefficients. In practical terms: the radical right coefficient is identified by years within a given country when its radical right vote share was above or below that country’s own average, controlling for years when all countries experienced common shocks. Countries that have not experienced meaningful variation in populism vote share over the period contribute little identifying power; the variation is primarily supplied by countries such as Hungary, Italy, France, Sweden, and Greece, whose radical right shares have moved substantially since 2001. 4.2 The Baseline Specification The baseline estimating equation follows the change specification of Celico and Rode ( 2023 ) and Rode and Revuelta ( 2015 ), augmented with EU membership: ΔEFW_it = β₁ EFW_{t-1} + β₂ POP_{t-1} + β₃ JC_{t-1} + β₄ logGDPpc_{t-1} + β₅ EU_{t-1} + δᴵ + γₜ + ε_it where ΔEFW_it is the annual change in the EFW overall index; EFW_{t-1} is the lagged EFW level capturing mean-reversion; POP_{t-1} is the lagged populism vote share; JC_{t-1} is the lagged Judicial Constraints index; logGDPpc_{t-1} is the lagged log GDP per capita; EU_{t-1} is the lagged EU membership indicator; and δᴵ, γₜ are country and year fixed effects respectively. Standard errors are clustered at the country level. 4.3 Ideology Decomposition and Interactions In Model 4, I replace aggregate POP with radical right (RRIGHT) and radical left (RLEFT) vote shares and add ideology-by-JC interactions: ΔEFW = β₁EFW_{t-1} + β₂RRIGHT_{t-1} + β₃RLEFT_{t-1} + β₄(RRIGHT×JC)_{t-1} + β₅(RLEFT×JC)_{t-1} + β₆JC_{t-1} + β₇logGDPpc_{t-1} + β₈EU_{t-1} + δᴵ + γₜ + ε The marginal effect of radical right populism is ∂(ΔEFW)/∂RRIGHT = β₂ + β₄ × JC. A positive β₄ suggests that stronger judicial constraints moderate (attenuate) the negative association. Note that β₂ is the coefficient when JC = 0, a hypothetical value far outside the observed sample range (minimum observed JC = 0.507); it should not be interpreted as the effect at any empirically plausible institutional level. The estimated marginal effects at observed JC quantiles are reported in Table 3 . 4.4 Sub-component Regressions I run five sub-component regressions replacing ΔEFW Overall with ΔEFW Area k (k = 1,...,5). Area-specific controls are included: log government consumption (% GDP) for Area 1, size of goverment; log(1 + CPI inflation/100) × 100 for Area 3, sound money; mean applied tariff rate for Area 4, freedom to trade. Areas 2 (legal system and property rights) and 5 (regulation) do not include area-specific raw controls because the principal sub-components of those areas (legal institution ratings, regulatory quality scores) are themselves the outcomes of interest rather than inputs available as separate regressors in the EFW master dataset. All specifications include EU membership and country and year fixed effects. 4.5 Robustness Three robustness checks are reported: (1) three-year rolling average ΔEFW as dependent variable to smooth annual EFW noise—I note that this creates a temporal mismatch since the populism regressor remains annual; a cleaner test would match three-year rolling averages of both variables, which is proposed for future work; (2) binary Populists in Government indicator instead of continuous vote share, which isolates the direct governance channel; (3) V-Dem Liberal Component Index (LCI) substituted for Judicial Constraints, as a more generic mechanism of institutional constraints, than include parliament and independents authorities. 4.6 Estimation, Inference, and Caveats Models are estimated using PanelOLS from the Python linearmodels 7.0 package with within-transformation for entity and time effects. Standard errors are clustered at the country level. With 31 clusters, conventional clustered SEs fall below the threshold of approximately 50 clusters recommended for reliable inference (Cameron, Gelbach and Miller 2008 ; MacKinnon 2019 ). Wild cluster bootstrap (WCB) SEs with 4,999 Rademacher draws (Roodman et al., 2019 ) are therefore implemented and reported in Table 7 as the preferred inference baseline. The Wooldridge (2002) test for first-order serial correlation in panel residuals yields rho = − 0.040 (H₀: rho = − 0.5; t = 9.38, p < 0.001), confirming the presence of AR(1) serial correlation. Country-clustered standard errors, which permit arbitrary within-country serial correlation without restricting the autocorrelation structure, are the appropriate response to this finding; the WCB provides additional robustness under the Rademacher distribution. The Arellano-Bond system-GMM estimator, which additionally addresses Nickell ( 1981 ) bias and the endogeneity of the populism regressor, remains the recommended extension for future work with a larger cross-section. One feature of the WCB results warrants clarification: the WCB standard error for the Model 1 radical right coefficient (0.00252) is marginally smaller than the conventional clustered SE (0.00260). This is not a reporting error, because with 31 clusters, the conventional asymptotic finite-sample correction can slightly over-inflate the variance estimate, while the Rademacher-draw bootstrap distribution, which is computed directly from within-cluster residuals, converges on a tighter empirical spread under certain configurations of within-cluster serial correlation. The WCB p-value (0.030) is arithmetically consistent with the marginally smaller bootstrap SE. I must note that despite the fact that several statistical tests have been applied to ensure as robust relationships as possible, the results should be interpreted as associations, since no causal model was employed. Two main endogeneity concerns apply: (1) reverse causality—declining EFW attracting populist voters—and (2) omitted time-varying confounders. As discussed in Section 6.4 , a panel Granger-causality test is implemented in this version. Bergh and Kärnä ( 2024 ) find no robust panel evidence that EFW changes predict populism vote share changes in European democracies; the extended Granger tests in Section 6.4 corroborate this finding within the sample across multiple lag structures. 5. Results 5.1 Main Models: Overall EFW Table 2 presents four nested models. Three patterns emerge consistently. First, the lagged EFW level is negative and highly significant (coefficient ≈ − 0.25, p < 0.001), confirming mean-reversion. Second, the Judicial Constraints index is positive and significant in Models 1 and 3 (p < 0.05 and p < 0.10), indicating that stronger judicial oversight correlates with subsequent economic freedom improvements. Third, log GDP per capita and EU membership are statistically insignificant throughout, reflecting limited within-country variation once country fixed effects are absorbed. Aggregate populism vote share in Model 1 has a coefficient of − 0.00038 (SE = 0.00051, p = 0.462)—small, negative, and indistinguishable from zero. This confirms the finding of Celico and Rode ( 2023 ) for OECD-type settings that aggregate populism does not significantly correlate with EFW changes in high-income institutionally constrained democracies. In Model 2, the interaction of populism vote share with Judicial Constraints is similarly negligible. In Model 3, both radical right (− 0.00076, p = 0.310) and radical left (− 0.00008, p = 0.922) vote shares are individually insignificant. Model 4, the most informative specification, includes ideology-by-JC interaction terms. The direct coefficient on radical right vote share is − 0.00444 (SE = 0.00206, p = 0.031 conventional; WCB p = 0.030, Table 7 )—statistically significant at the 5 per cent level under both conventional and bootstrap inference. The interaction RRIGHT × JC is positive and marginally significant (0.00419, SE = 0.00227, p = 0.065). The marginal effect of radical right populism is therefore: ∂(ΔEFW)/∂RRIGHT = − 0.00444 + 0.00419 × JC. At the sample mean (JC = 0.902), the marginal effect is − 0.00060 per percentage point of vote share; multiplied by one standard deviation (13.5 pp), this implies − 0.008 EFW points per year, or approximately − 0.03 EFW points over a four-year electoral cycle. The radical left coefficient in Model 4 is positive and not significant (0.01764, p = 0.124). For the radical left null result, I note that the observed mean radical left vote share is 6.6 per cent (SD = 9.5 pp)—substantially lower prevalence than radical right in this European sample. Under the Model 4 specification, a one-standard-deviation increase in radical left vote share (9.5 pp) would imply an effect of only 0.00167 EFW points per year at the sample mean JC—an economically negligible magnitude even if statistically non-zero. The null result thus reflects both a baseline-shift mechanism (discussed in Section 6.2 ) and insufficient statistical power given lower prevalence and variance of radical left vote shares. One caveat merits acknowledgment: the Rad. Left × JC interaction term, while not significant under conventional clustered inference (Table 2 ), yields WCB p = 0.054 (Table 7 )—borderline at the 10 per cent level. This does not alter the substantive interpretation—the radical left coefficient on ΔEFW remains economically negligible at all empirically plausible JC values, however, the marginal bootstrap result warrants cautious acknowledgment rather than a conclusive null finding. Figure 1 visualises the marginal effects of the main finding, that of the radical right populism on ΔEFW at selected JC levels. Table 2 Main Regression Results: Annual Change in EFW Overall Variable M1 M2 M3 M4 EFW (t − 1) −0.2491*** −0.2493*** −0.2559*** −0.2559*** (0.0302) (0.0313) (0.0316) (0.0321) Populism Vote Share (t − 1) −0.0004 −0.0001 (0.0005) (0.0035) Pop × Judicial Constraints (t − 1) −0.0002 (0.0039) Radical Right Vote Share (t − 1) −0.0008 −0.0044** (0.0007) (0.0021) Radical Left Vote Share (t − 1) −0.0001 0.0176 (0.0009) (0.0115) Rad. Right × Judicial Constraints (t − 1) 0.0042* (0.0023) Rad. Left × Judicial Constraints (t − 1) −0.0203 (0.0135) Log GDP per capita (t − 1) −0.0188 −0.0188 −0.0111 −0.0017 (0.0520) (0.0520) (0.0509) (0.0551) Judicial Constraints (t − 1) 0.1436** 0.1498 0.0931* −0.0150 (0.0630) (0.1752) (0.0537) (0.0989) EU Membership (t − 1) 0.0096 0.0097 0.0068 0.0096 (0.0225) (0.0225) (0.0213) (0.0225) Observations 712 712 663 663 Within-R² 0.254 0.254 0.233 0.237 Country FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes Clusters 31 31 31 31 Note: Dependent variable: annual change in EFW overall index. All regressors lagged one period. Country-clustered standard errors in parentheses. *** p < 0.01; ** p < 0.05; * p < 0.10. PanelOLS with entity and time fixed effects. M3–4 use 663 obs. due to availability of radical right/left decomposition in TAP. The M3 aggregate null for radical right and left is consistent with the M1 result, suggesting sample attrition does not drive findings. Table 3 Marginal Effects of Radical Right Populism on ΔEFW (Model 4) JC Percentile / Value JC Value Marginal Effect (per pp) 1 SD Effect (×13.5 pp) 10th percentile (institutionally weaker) 0.792 −0.00107 −0.0144 25th percentile 0.863 −0.00074 −0.0100 Sample mean 0.902 −0.00060 −0.0082 75th percentile 0.957 −0.00036 −0.0049 90th percentile (institutionally stronger) 0.978 −0.00029 −0.0039 Crossover (ME = 0) 1.047 (out of sample) 0.00000 — Note: Marginal effect = ∂(ΔEFW)/∂(Rad. Right Vote Share) = − 0.00444 + 0.00419 × JC (Model 4). “1 SD Effect” is the change in ΔEFW associated with a one SD increase in radical right vote share (13.5 pp). The crossover at JC = 1.047 lies outside the observed sample range (max = 0.988), confirming the negative association is present at all empirically plausible JC levels. 5.2 Sub-component Regressions Table 4 presents sub-component regressions using aggregate populism vote share. No sub-component shows a statistically significant association with aggregate populism vote share at conventional levels. Coefficients are uniformly small, ranging from − 0.00252 (Area 3: Sound Money, p = 0.175) to + 0.00080 (Area 5: Regulation, p = 0.485). In Table 4 , the Judicial Constraints index is significant for Areas 2 and 3 (p < 0.05 each), confirming that stronger judicial oversight is associated with improvements in both the legal system and sound money sub-components, independent of populism. The EU membership variable is not significant for any sub-component, though the directional coefficient for Area 3 is positive (0.109, p = 0.187). In the ideology decomposition (Table 5 ), Area 4 (Freedom to Trade Internationally) exhibits a negative association with radical right populism vote share (coefficient = − 0.00251, SE = 0.00119, conventional p = 0.041). Under the preferred wild cluster bootstrap inference (Table 7 ), the WCB p-value is 0.063—marginally significant at the 10 per cent level but not crossing the conventional 5 per cent threshold. The conventional p-value without the EU membership control is 0.070. WCB p = 0.063 should be treated as the definitive estimate for this coefficient; the conventional p = 0.041 is reported for comparability with specifications that do not implement bootstrap inference. Area 1 (Government Size) shows a negative but insignificant coefficient for radical left populism (− 0.00330, p = 0.247), consistent in direction with left-populist expansion of government spending, however, the power of this relationship is insufficient for a firm conclusion. Figure 2 visualises the radical right and radical left populism coefficients across EFW sub-components (Table 5 specification). Table 4 Sub-component Regressions: Aggregate Populism Vote Share Variable Area 1 Gov. Size Area 2 Legal/Prop. Area 3 Snd. Money Area 4 Trade Area 5 Regulation EFW Area (t − 1) −0.188*** −0.308*** −0.135*** −0.206*** −0.182*** (0.039) (0.045) (0.032) (0.035) (0.036) Populism Vote Share (t − 1) −0.0019 0.0008 −0.0025 0.0002 0.0008 (0.0017) (0.0007) (0.0018) (0.0009) (0.0011) Judicial Constraints (t − 1) 0.103 0.249** 0.420** 0.042 0.009 (0.147) (0.116) (0.177) (0.108) (0.100) EU Membership (t − 1) 0.033 −0.003 0.109 0.037 0.048 (0.040) (0.021) (0.084) (0.054) (0.031) Log GDP p.c. (t − 1) −0.055 0.053 0.063 −0.018 0.003 (0.150) (0.062) (0.109) (0.067) (0.073) Area-specific control Log govt cons. No Log(1 + infl) No No Observations 712 712 712 712 712 Within-R² 0.105 0.156 0.061 0.099 0.052 Country & Year FE Yes Yes Yes Yes Yes Note: Dependent variables: annual changes in EFW sub-areas 1–5. All regressors lagged one period. Country-clustered SEs in parentheses. *** p < 0.01; ** p < 0.05; * p < 0.10. Table 5 Sub-component Regressions: Ideology Decomposition Variable Area 1 Gov. Size Area 2 Legal/Prop. Area 3 Snd. Money Area 4 Trade Area 5 Regulation Radical Right Vote Share (t − 1) −0.0009 −0.0003 −0.0015 −0.0025** 0.0009 (0.0022) (0.0009) (0.0022) (0.0012) (0.0010) Radical Left Vote Share (t − 1) −0.0033 0.0010 −0.0005 −0.0002 0.0015 (0.0029) (0.0009) (0.0041) (0.0010) (0.0028) EFW Area (t − 1) −0.188*** −0.289*** −0.131*** −0.202*** −0.172*** (0.040) (0.046) (0.032) (0.036) (0.038) EU Membership (t − 1) 0.032 −0.003 0.105 0.025 0.046 (0.040) (0.022) (0.085) (0.055) (0.031) Observations 663 663 663 663 663 Within-R² 0.097 0.140 0.057 0.095 0.051 Country & Year FE Yes Yes Yes Yes Yes Note: Dependent variables: annual changes in EFW sub-areas 1–5. Radical right and left vote shares included simultaneously. Country-clustered SEs in parentheses. *** p < 0.01; ** p < 0.05; * p < 0.10. The Area 4 radical right coefficient (p = 0.041) is borderline: p = 0.070 without EU membership control. JC coefficient suppressed for brevity; full results available upon request. Table 6 Robustness Checks Specification Coeff. on Populism SE p-value Baseline (M1): Annual ΔEFW, vote share, EU control −0.00038 0.00051 0.462 Robustness 1: 3-yr rolling avg ΔEFW, vote share¹ −0.00019 0.00054 0.689 Robustness 2: Annual ΔEFW, binary populist-in-gov −0.00932 0.01030 0.377 Robustness 3: LCI instead of JC as institutional proxy −0.00026 0.00051 0.610 Note: All models include country and year FEs, EU membership dummy, clustered SEs (31 clusters). ¹Rolling average creates temporal mismatch (annual populism vs. smoothed EFW); a matched rolling-average populism variable would constitute a cleaner test. Table 7 Wild Cluster Bootstrap Inference (4,999 Rademacher Draws) Variable / Model Coeff. Clustered SE Clust. p WCB SE WCB p M1: Rad. Right Vote Share −0.00484 0.00260 0.063 0.00252 0.030** M1: Rad. Left Vote Share −0.00384 0.00227 0.091 0.00215 0.054+ M4: Rad. Right × JC −0.01827 0.01040 0.080 0.00988 0.065+ M4: Rad. Left Vote Share 0.03409 0.02315 0.141 0.02201 0.130 M4: Rad. Left × JC −0.04349 0.02472 0.079 0.02351 0.054+ Area 4: Rad. Right Vote Share −0.00609 0.00360 0.093 0.00343 0.063+ Area 4: Rad. Left Vote Share −0.00592 0.00406 0.147 0.00406 0.152 Note: Wild cluster bootstrap via double-demeaned within-transformation with Rademacher cluster-sign draws (B = 4,999, seed = 42). WCB p-values are two-sided. ** WCB p < 0.05; + WCB p < 0.10. 5.3 Backsliding versus Stable-Democracy Heterogeneity Section 4.1 notes that the JC×RRIGHT interaction in Model 4 is primarily identified from countries experiencing democratic backsliding— that is Hungary (JC decline post-2010) and Poland (JC decline post-2015). Hungary’s V-Dem Judicial Constraints index fell from approximately 0.92 in 2010 to 0.78 by 2020, a decline of roughly 14 percentage points on the 0–1 scale; Poland’s index declined from approximately 0.90 in 2015 to below 0.80 by 2021. Both countries are placed well below the sample mean of 0.903 by the end of the backsliding period, providing the primary within-country variation that identifies the backsliding interaction. To assess whether the radical right–EFW relationship differs systematically between backsliding and stable democracies, I augment the Model 1 specification with a backsliding dummy (BS = 1 for Hungary from 2010, Poland from 2015, and 0 otherwise) and its interactions with radical right and radical left vote shares. The results are reported in Table 8 . The base radical right coefficient (effect in stable democracies) is − 0.00519 (SE = 0.00306, p = 0.090), directionally consistent with Model 1 and marginally significant. The interaction RRIGHT×BS is positive and statistically significant at the 5 per cent level (0.00649, SE = 0.00269, p = 0.016), indicating that the negative association between radical right vote share and EFW changes is substantially attenuated during backsliding episodes. The combined marginal effect during backsliding periods is approximately zero (+ 0.00130). This pattern is consistent with the constitutional capture mechanism: as Orbán’s Fidesz and Kaczyński’s PiS consolidated power, they selectively liberalised business regulation and pursued supply-side tax reductions—policies that improve EFW Area 1 and Area 5 scores—while simultaneously eroding judicial independence and trade openness. The near-zero net effect during backsliding reflects this mixed institutional legacy: democratic regression (negative for Area 2) partly offset by business-friendly economic deregulation (positive for Areas 1 and 5). The radical left interaction RLEFT×BS is negative but imprecisely estimated (− 0.155, SE = 0.093, p = 0.097). Table 8 Backsliding Heterogeneity: Radical Right vs. Stable Democracy Split Variable Coeff. SE p-value Rad. Right Vote Share (stable democracies) −0.00519 0.00306 0.090+ Rad. Left Vote Share (stable democracies) −0.00393 0.00227 0.084+ Backsliding Dummy (HUN 2010+, POL 2015+) 0.00650 0.00318 0.042** Rad. Right × Backsliding 0.00649 0.00269 0.016** Rad. Left × Backsliding −0.15469 0.09297 0.097+ ME of Rad. Right during backsliding 0.00130 — — Observations 713 Country & Year FE Yes Note: Dependent variable: annual ΔEFW overall. Backsliding dummy = 1 for Hungary ≥ 2010 and Poland ≥ 2015, 0 otherwise. ME during backsliding = base RR coeff + RR×BS coeff = − 0.00519 + 0.00649 = + 0.00130. Country-clustered SEs. ** p < 0.05; + p < 0.10. 6. Discussion 6.1 The European Panel as a Demanding Test The results contribute to the literature in a specific and bounded way. The null result for aggregate populism in European democracies can be interpreted as precisely what was expected by the theoretical framework and what Celico and Rode ( 2023 ) find for OECD settings. European democracies in the sample have high and relatively stable Judicial Constraints scores (mean JC = 0.903), and the institutional safeguards present in most of these polities appear sufficient to absorb or deflect, at least, short-run populist-driven policy changes. The homogeneity of the sample, which is a design strength by choice for internal validity, simultaneously suggests attenuation bias: the true coefficient on radical right populism in a more institutionally diverse global sample is likely larger in absolute value than the conservative estimates presented here. I note the significance of the crossover JC value (1.047, outside the sample range) with more nuance than a simple report might convey. Its implication is not that judicial constraints are irrelevant, they demonstrably attenuate the negative association as the interaction coefficient shows, but rather that no European democracy in the sample sits at a JC level high enough to fully neutralise the negative association of radical right populism. At the 90th percentile of JC (0.978), the 1 SD annual effect is only − 0.004 EFW points: statistically non-zero at current coefficient estimates but economically negligible. The appropriate conclusion is therefore: stronger judicial institutions systematically attenuate the association, but even the institutionally strongest European democracies do not eliminate it entirely. 6.2 Trade Freedom as the Most Susceptible Sub-component The significant negative association between radical right populism and Area 4 (Trade Freedom, WCB p = 0.063; conventional p = 0.041 with EU membership control, p = 0.070 without) is the strongest sub-component finding. This result requires a theoretical reconciliation with Bergh and Kärnä’s ( 2024 ) emphasis on Area 2 (rule of law) as the central nexus between economic freedom and right-wing populism. The two findings are not contradictory but operate in different directions and timeframes: Bergh and Kärnä find that countries with weaker rule of law produce more right-wing populism (the reverse direction, using EFW levels), while the results here show that rising radical right vote shares are associated with deteriorations in trade freedom (the direction of this paper, using EFW changes). Trade Freedom is the sub-component most directly actionable through electoral and governmental channels: tariff rates, non-tariff barriers, and regulatory trade restrictions are standard legislative instruments that governments facing populist pressure may deploy in the short run, without necessarily requiring the constitutional or judicial modifications needed to erode Area 2. The public choice mechanism can be instructive here: protectionism serves as a coalition-building instrument for radical right parties, directing concentrated benefits to politically organised domestic industries (manufacturing, agriculture) while dispersing costs over unorganised consumers and trading partners. This mechanism operates both through direct policy when populists govern and through incumbents’ anticipatory concessions to electoral pressure—explaining why the vote-share specification captures a significant signal even absent full government participation. The null Area 2 result in this specification does not contradict Bergh and Kärnä; it reflects the short-run nature of this annual change specification and the difficulty of capturing gradual institutional erosion in a first-differenced framework. The empirical literature provides strong direct support for this protectionist characterisation of European radical right parties. Colantone and Stanig ( 2018 ) show that regional exposure to Chinese import competition across 15 Western European countries significantly increased vote shares for nationalist and right-wing populist parties—and that these parties subsequently adopted explicitly isolationist and protectionist platforms in response to their electorate’s trade-displaced support base. Dippel, Gold, Heblich and Pinto ( 2019 ) replicate this mechanism for Germany, tracing the path from import exposure to far-right votes via xenophobic attitudes. Rodrik ( 2018 ) provides the unifying theoretical account: European right-wing populism is fundamentally a form of economic nationalism, combining hostility to trade openness with demands for national industrial protection, and operating through the specific channel of tariffs, non-tariff barriers, and resistance to multilateral trade agreements—precisely the components measured by EFW Area 4. The null finding for radical left populism in Area 4 is best understood through a baseline-shift argument rather than simply as a statistical failure. Left parties and their mainstream centre-left counterparts were already substantially converged on anti-trade, pro-intervention platforms well before the populist surge of the 2010s: the incumbent centre-left had been absorbing redistributive, trade-sceptical, and capital-control preferences into its programme for decades, although a part of the party family during 1990s (the third way) saw liberalization as a positive policy choice. A surge in radical left vote share therefore creates little measurable policy distance—the counterfactual starting point is already close to the populist position. The radical right, by contrast, represented a genuine ideological rupture from a centre-right tradition that had been broadly pro-trade and pro-globalisation since the Thatcher-CDU-Maastricht consensus of the 1980s and 1990s (Inglehart and Norris 2016 ; Rodrik 2018 ). The shift of the right coalition toward protectionism was discontinuous and electorally driven, generating measurable policy distance and thus a detectable Area 4 signal in the data. This is consistent with the party competition literature on incumbent responsiveness to flanking parties: incumbents adjust platforms in proportion to the ideological distance they must travel (Adams et al., 2004 ; Somer-Topcu 2009 ). A secondary mechanism reinforces the left null: the most prominent left-populist governments in the sample, that is SYRIZA-ANEL in Greece (2015–2019) and the Podemos-supported coalition in Spain, governed under EU fiscal surveillance, Eurozone membership, and, in the Greek case, full Troika conditionality. These external constraints effectively compressed the policy space available to left populists in power, forcing a degree of programme moderation that their electoral mandate did not reflect. SYRIZA under Tsipras ultimately signed austerity measures that were in complete opposite direction of its electoral programme a few months earlier; similarly, Podemos in coalition was constrained by EU deficit rules. This ‘moderation in office’ dynamic is structurally asymmetric: Orbán’s Hungary and Kaczyński’s Poland were net EU budget recipients facing no equivalent external fiscal disciplining mechanism, and Article 7 proceedings against them required unanimity—a far weaker constraint than Troika conditionality. The right populist programmes could therefore be partially implemented, while the left populist programmes largely could not. This asymmetry in external constraint, rather than any inherent moderation of left populism, accounts in part for the differential signals in the data. 6.3 Comparison with Celico and Rode ( 2023 ) and Bergh and Kärnä ( 2024 ) This study was primarily initiated to discuss with the results of the tow prior studies. The findings align with Celico and Rode ( 2023 ) on the aggregate null for OECD democracies, while seeking to provide a more granular decomposition on both ideology and EFW components. Their finding that the association between populism and EFW in OECD settings is mediated by political ideology is corroborated by the ideology decomposition: the null aggregate result in Models 1–3 masks a conditional negative direct effect of radical right populism that is moderated by judicial constraints (Model 4). The main difference is that the results show a consistently null result for left-wing populism in the European panel—likely because left-populist episodes (SYRIZA, Podemos) are less prevalent and shorter-lived in the sample than the historically left-dominant populism that drove Rode and Revuelta’s ( 2015 ) results, and because of the external constraint asymmetry described in Section 6.2 . As discussed in Section 6.2 above, While Bergh and Kärnä ( 2024 ) identify weak rule of law as the structural condition breeding right-wing populism, this study finds that once populists gain electoral ground, it is trade freedom — the most legislatively accessible dimension of economic freedom — that deteriorates first. 6.4 Limitations Several limitations must be acknowledged. First, all results reflect correlational relationships with no causal effects mechanisms, by design. The potential for reverse causality, that is declining economic freedom contributing to populist electoral success, cannot be fully excluded by lagged regressors alone. Although Bergh and Kärnä ( 2024 ) find no robust panel evidence that changes in aggregate economic freedom predict changes in populism vote shares across European democracies, I additionally implement a Dumitrescu–Hurlin (2012) panel Granger-causality test—asking whether lags of radical right (or radical left) vote share predict EFW changes, country-by-country, with the average F-statistic aggregated into a W-statistic—across three lag structures (2, 3, and 5 years) to assess robustness at short to medium-run horizons. Table 9 reports the results. Across all lag structures and for both vote-share variables, the test consistently fails to reject H₀ of no panel Granger causality (all p > 0.90). These results strongly corroborate the signalling-channel interpretation that populist vote shares do not lead EFW changes, at least, in the short to medium run; any causal pathway would require a longer horizon than five-year lags can detect, since huge institutional changes require a long term erosion; maybe Hungary’s 16 years case could provide such a institutional ground for further research. In addition, a credible structural causal identification strategy—instrumental variables or regression discontinuity around close populist election victories—remains an important agenda for future work. Table 9 Extended Dumitrescu–Hurlin Panel Granger Causality Tests Lags Variable avg-F W p-value G (countries) N (obs) 2 Radical Right Vote Share 1.445 44.8 0.951 31 620 2 Radical Left Vote Share 0.890 26.7 > 0.99 30 600 3 Radical Right Vote Share 1.271 39.4 > 0.99 31 589 3 Radical Left Vote Share 0.800 24.8 > 0.99 31 589 5 Radical Right Vote Share 1.659 51.4 > 0.99 31 527 5 Radical Left Vote Share 1.021 31.6 > 0.99 31 527 Note: Dumitrescu–Hurlin (2012) panel Granger test. H₀: populism vote shares do not Granger-cause ΔEFW. W = sum of country-level F-statistics; p-value based on chi²(G×lags) approximation. G = number of countries with sufficient observations. All p-values well above 0.10 across all lag structures, confirming the absence of a short-to-medium-run reverse-causality channel. Second, the sample of 31 European countries is geographically restricted, by selection design. Findings should not be generalised to Latin American, African, or Asian democracies where the institutional and ideological character of populism differs substantially. Third, with 31 country clusters, conventional clustered standard errors may be insufficiently conservative; wild cluster bootstrap inference is implemented in Table 7 and generally confirms the directional results, with the radical right coefficient in Model 1 strengthening to WCB p = 0.030. Fourth, the within-term constancy of populism vote share—carried forward from election years—limits annual within-term variation; the robustness checks suggest this does not drive the results. Fifth, the Nickell ( 1981 ) bias from the lagged dependent variable is expected to be negligible at T ≈ 23 years, but an Arellano-Bond GMM estimator remains the recommended extension. Sixth, non-electoral channels of populist influence—referendum campaigns (Brexit), executive decrees, and supra-parliamentary mobilisation—are not captured by vote-share or government-participation indicators. Seventh, the JC×RRIGHT interaction is primarily identified from countries experiencing democratic backsliding (Hungary, Poland); Section 5.3 presents the backsliding heterogeneity analysis, finding that the negative radical right–EFW association is present in stable democracies (p = 0.090) and attenuated in backsliding episodes. Eighth, the Wooldridge (2002) test confirms first-order serial correlation in panel residuals (rho = − 0.040 vs. H₀: rho = − 0.5; t = 9.38, p < 0.001). Country-clustered standard errors and the wild cluster bootstrap address this. On the question of structural identification an ideal instrument for radical right party strength would require variation in the supply of populist candidates or party funding that is exogenous to both domestic economic conditions and EFW trajectories. The strategic foreign funding of intellectual and political infrastructure by sympathetic foreign governments—channelled through foundations and non-governmental organizations indirectly, rather than direct party finance, represents one possible identification structure; nevertheless, the endogenous targeting of already-receptive environments and the opacity of financial flows would likely undermine the exclusion restriction in practice. 7. Conclusion This paper has examined the association between populism and economic freedom across 31 European democracies from 2001 to 2023, extending the framework of Celico and Rode ( 2023 ) and complementing the reverse-direction analysis of Bergh and Kärnä ( 2024 ). I introduced three novel dimensions: simultaneous ideology decomposition (radical right vs. radical left), sub-component regressions for all five EFW areas, and a backsliding heterogeneity analysis distinguishing institutionally stable democracies from democratic backsliders. EU membership was included as an additional time-varying control, while the signalling versus policy channel distinction was made explicit in the theoretical framework. Aggregate populism vote share is not significantly associated with annual EFW changes, confirming the institutional mediation prediction for OECD-type democracies with high institutional framework. When decomposed by ideology, radical right populism has a significant direct negative association attenuated by judicial constraint strength—economically small at the sample mean but potentially cumulative over electoral cycles. The sub-component analysis identifies Freedom to Trade Internationally as the EFW component most sensitive to radical right populism (marginally significant at the 10 per cent level under WCB inference: WCB p = 0.063; conventional p = 0.041), consistent with the protectionist orientation of contemporary European radical right platforms and the public choice logic of protectionism as coalition-building. The radical left null result is consistent with a baseline-shift mechanism—left parties were already converged on anti-trade positions, leaving no measurable policy distance to cover—reinforced by the external constraints that disciplined left-populist governments in practice (Troika conditionality, Eurozone fiscal rules), an asymmetry not shared by the right populist governments in Hungary and Poland. The paper contributes a focused European test, an explicit side-by-side comparison of the two most closely related predecessor studies, and the first sub-component analysis in this literature to identify which EFW channels are most sensitive to populist electoral strength. Future research priorities should employ (1) Arellano-Bond GMM robustness and stronger causal identification through electoral quasi-experiments; (2) a global panel extension; (3) longer-horizon effects on EFW legal system and sound money areas, which theory suggests are more vulnerable to sustained populist governance than to short-run electoral pressure; (4) party-level analysis to disaggregate within-family heterogeneity in economic programmes; and (5) a matched rolling-average specification pairing populism and EFW changes at the same temporal frequency. Declarations Competing Interests The author declares no competing interests, financial or non-financial, related to the work submitted for publication. Author Contribution Constantinos Saravakos conceived and designed the study, collected and processed the data, conducted all empirical analyses, interpreted the results, and wrote the manuscript in its entirety. Acknowledgement The author thanks Robert Lawson (Jerome M. Fullinwider Centennial Chair in Economic Freedom and Director of the Bridwell Institute for Economic Freedom, Cox School of Business, Southern Methodist University), Dr. Dan Mitchell (Center for Freedom and Prosperity), and Ioannis Filios for helpful comments and suggestions on earlier drafts of this manuscript. The usual disclaimer applies. Data Availability The data that support the findings of this study are drawn from four publicly available sources. Economic freedom data are from the Fraser Institute Economic Freedom of the World 2025 Annual Report and Master Index Dataset, available at https://www.fraserinstitute.org/economic-freedom/dataset. Populism vote share data are from Timbro Authoritarian Populism Index the https://populismindex.com/. Institutional data are from the Varieties of Democracy (V-Dem) Dataset v16, available at https://www.v-dem.net/data/the-v-dem-dataset. Macroeconomic controls are from the World Bank World Development Indicators, available at https://databank.worldbank.org/source/world-development-indicators. Python code for the wild cluster bootstrap and Dumitrescu–Hurlin panel Granger tests was developed with AI-assisted technical support (Claude, Anthropic). References Absher S, Grier K, Grier R (2020) The economic consequences of durable left-populist regimes in Latin America. J Econ Behav Organ 177:787–817 Adams J, Clark M, Ezrow L, Glasgow G (2004) Understanding change and stability in party ideologies: Do parties respond to public opinion or to past election results? Br J Polit Sci 34(4):589–610 Autor D, Dorn D, Hanson G, Majlesi K (2020) Importing political polarization? The electoral consequences of rising trade exposure. Am Econ Rev 110(10):3139–3183 Bennett DL, Boudreaux C, Nikolaev B (2021) Populist discourse and entrepreneurship: The role of political ideology and institutions. J Int Bus Stud 52(4):800–810 Bennett DL, Boudreaux C, Nikolaev B (2023) Populism and economic freedom. J Comp Econ 51(2):574–597 Berggren N (1999) Economic freedom and equality: friends or foes? Public Choice 100(3):203–223 Bergh A, Kärnä A (2024) Economic freedom and populism. In: Berggren N (ed) Handbook of Research on Economic Freedom. Edward Elgar, Cheltenham, pp 297–314 Bermeo N (2016) On democratic backsliding. J Democr 27(1):5–19 Bjørnskov C, Potrafke N (2012) Political ideology and economic freedom across Canadian provinces. East Econ J 38(2):143–166 Bjørnskov C, Potrafke N (2013) The size and scope of government in the US states: does party ideology matter? Int Tax Public Finance 20(4):687–714 Bjørnskov C, Rode M (2020) Regime types and regime change: A new dataset on democracy, coups, and political institutions. Rev Int Organ 15(2):531–551 Cameron AC, Gelbach JB, Miller DL (2008) Bootstrap-based improvements for inference with clustered errors. Rev Econ Stat 90(3):414–427 Celico A, Rode M (2023) Populism, majority rule, and economic freedom. In: Gwartney J, Lawson R, Murphy R (eds) Economic Freedom of the World: 2023 Annual Report. Fraser Institute, Vancouver Celico A, Rode M, Rodríguez Carreño J (2022) Random forest populism indicators. Working Paper, Universidad de Navarra Colantone I, Stanig P (2018) The trade origins of economic nationalism: Import competition and voting behavior in Western Europe. Am J Polit Sci 62(4):936–953 Coppedge M et al (2023) V-Dem [Country-Year/Country-Date] Dataset v13. Varieties of Democracy (V-Dem) Project de Haan J, Sturm J-E (2000) On the relationship between economic freedom and economic growth. Eur J Polit Econ 16(2):215–241 Dippel C, Gold R, Heblich S, Pinto R (2019) The path from trade to right-wing populism in Europe. Elect Stud 60:102040 Dumitrescu E-I, Hurlin C (2012) Testing for Granger non-causality in heterogeneous panels. Econ Model 29(4):1450–1460 Funke M, Schularick M, Trebesch C (2023) Populist leaders and the economy. Am Econ Rev 113(12):3249–3288 Guriev S, Papaioannou E (2022) The political economy of populism. J Econ Lit 60(3):753–832 Gwartney J, Lawson R, Hall J, Murphy R (2023) Economic Freedom of the World: 2023 Annual Report. Fraser Institute, Vancouver Gwartney J, Lawson R, Holcombe R (1999) Economic freedom and the environment for economic growth. J Inst Theor Econ 155(4):643–663 Hall JC, Lawson RA (2014) Economic freedom of the world: An accounting of the literature. Contemp Econ Policy 32(1):1–19 Halikiopoulou D, Vlandas T (2019) What is new and what is nationalist about Europe’s new nationalism? Explaining the rise of the far right in Europe. Nations Natl 25(2):409–434 Heinö AJ (2024) Authoritarian Populism Index 2024. Timbro, Stockholm. www.populismindex.com Henisz WJ (2000) The institutional environment for economic growth. Econ Polit 12(1):1–31 Huber RA, Schimpf CH (2017) On the distinct effects of left-wing and right-wing populism on democratic quality. Polit Gov 5(4):146–165 Inglehart R, Norris P (2016) Trump, Brexit, and the rise of populism: Economic have-nots and cultural backlash. HKS Working Paper RWP16-026. Harvard Kennedy School, Cambridge Jäger K (2017) Economic freedom in the early 21st century: Government ideology still matters. Kyklos 70(2):257–278 Lawson R (2022) Economic freedom of the world: Annual update. Indep Rev 27(1):3–12 Levitsky S, Ziblatt D (2018) How Democracies Die. Crown, New York Lührmann A et al (2020) Varieties of Party Identity and Organization (V-Party) Dataset V1. Varieties of Democracy. V-Dem) Project MacKinnon JG (2019) How cluster-robust inference is changing applied econometrics. Can J Econ 52(3):851–881 Margalit Y (2019) Economic insecurity and the causes of populism, reconsidered. J Econ Perspect 33(4):152–170 Meijers MJ, Zaslove A (2021) Measuring populism in political parties: Appraisal of a new approach. Comp Polit Stud 54(2):372–407 Mudde C, Rovira Kaltwasser C (2017) Populism: A Very Short Introduction. Oxford University Press, Oxford Nickell S (1981) Biases in dynamic models with fixed effects. Econometrica 49(6):1417–1426 Pitlik H (2007) A race to liberalization? Diffusion of economic policy reform among OECD-economies. Public Choice 132(1):159–178 Rode M, Gwartney J (2012) Does liberalization cause growth? An empirical investigation using EFW data. J Priv Enterp 27(2):1–28 Rode M, Revuelta J (2015) The wild bunch! An empirical note on populism and economic institutions. Econ Gov 16(1):73–96 Rodrik D (2018) Populism and the economics of globalization. J Int Bus Policy 1(1–2):12–33 Roodman D, Nielsen MØ, MacKinnon JG, Webb MD (2019) Fast and wild: Bootstrap inference in Stata using boottest. Stata J 19(1):4–60 Rooduijn M et al (2019) The PopuList: An overview of populist, far right, far left and Eurosceptic parties in Europe. Somer-Topcu Z (2009) Timely decisions: The effects of past national elections on party policy change. J Polit 71(1):238–248 Stöckl T, Rode M (2021) All that glitters: Populism, political risk, and the IMF. Kyklos 74(1):50–76 Team Populism (2024) TAP Dataset: Populist, Radical Right and Radical Left Parties in European Democracies. Brigham Young University, Provo Weyland K (1999) Neoliberal populism in Latin America and Eastern Europe. Comp Polit 31(4):379–401 Footnotes Throughout this paper, I employ the term ‘far-right’ as a definition for these political formations, following Halikiopoulou and Vlandas ( 2019 ), who demonstrate that ‘far-right’ captures both the populist and the nationalist-authoritarian dimensions that are jointly operative in parties such as Fidesz, PiS, and Lega. In tables and variable labels I retain ‘radical right’ to match the TAP 2024 dataset coding, which uses that term for the relevant party family. However, I note that no formal cross-dataset correlation exercise is conducted here. A harmonised polling series for 31 countries over 23 years is not currently available; the robustness checks (binary government-participation indicator, three-year rolling EFW average) provide partial validation that the carry-forward convention does not drive the results. Additional Declarations No competing interests reported. Supplementary Files Appendix.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9451071","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627291312,"identity":"cadb2cbf-e1e2-44cc-bc68-e1fd298a1be8","order_by":0,"name":"Constantinos Saravakos","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIie3PMQuCQBTA8ScHTprrHUZ9hZODIJA+ixE02dQiNNTkFq36LYLgZsHB5aDV1aWlhiQIgoaMmgpO3SLuD4+74X4PDkCl+sHw+6QAKHldk+ZE91oTgzYjZJMezrdwxCxLXHHpQ6+Te+gYSIiNpyxeh5MBiWacxBwYyT19KCSkh4GBIZBLc5PbJofxtiLOSkas7KLdxdKle3F4kmUtscFnyAjSAU18/Uk8Wv2lkBES+XPUDTJGoikbxhw7sShCTUbwPttpJ7pwNlZa5CV3+51skpYy8rWiGi3Ete8+Q+fWRKVSqf65B6nuSnZhjFNqAAAAAElFTkSuQmCC","orcid":"","institution":"University of Macedonia","correspondingAuthor":true,"prefix":"","firstName":"Constantinos","middleName":"","lastName":"Saravakos","suffix":""}],"badges":[],"createdAt":"2026-04-17 15:55:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9451071/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9451071/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107914692,"identity":"02a7080a-7272-415f-8f20-dbd10494549c","added_by":"auto","created_at":"2026-04-27 13:59:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":129337,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eMarginal Effects of Radical Right Populism on ΔEFW at Selected Judicial Constraint Levels (shaded area represents 95% delta-method confidence interval)\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9451071/v1/5a90bfc5c20069cf4c303384.png"},{"id":107914693,"identity":"e4d206f5-bc51-4f6f-b561-69b2283a467d","added_by":"auto","created_at":"2026-04-27 13:59:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":104495,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eRadical Right and Radical Left Populism Coefficients across EFW Sub-components (Table 5 specification)\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9451071/v1/2bf52212225a6f4fa7c8959c.png"},{"id":109081243,"identity":"8d6082d0-fef9-4cc3-8e78-c85851e92545","added_by":"auto","created_at":"2026-05-12 12:11:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":874733,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9451071/v1/5600e2c7-d5f5-419d-84c8-f2a34d19fce6.pdf"},{"id":107914691,"identity":"e6caf4b5-6c57-4bb6-9851-332f57138498","added_by":"auto","created_at":"2026-04-27 13:59:41","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15999,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-9451071/v1/be72b9cbde634b105c4dfdc8.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Populism and Economic Freedom in European Democracies: Evidence from a Two-Way Fixed Effects Panel Analysis, 2001–2023","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe past two decades have witnessed a sustained rise of populist parties across European democracies, transforming them from a marginal political phenomenon into a defining feature of electoral competition. From Hungary\u0026rsquo;s Fidesz to Italy\u0026rsquo;s Lega, Poland\u0026rsquo;s Law and Justice (PiS), and Sweden\u0026rsquo;s Democrats, far-right formations have managed to expand their vote shares substantially. At the same time, left-populist formations\u0026mdash;from Syriza in Greece to Podemos in Spain\u0026mdash;have gained significant electoral support, particularly following the 2008\u0026ndash;2010 financial crisis, but for a shorter period of time. By 2023, the mean populism vote share across the 31 European democracies in the sample had reached approximately 25 per cent, nearly double the level observed in 2001. This growth is primarily driven by far-right parties, whose aggregate vote shares have grown from roughly 7 per cent in 2001 to over 17 per cent by 2023, while radical left shares have remained more stable at approximately 6\u0026ndash;8 per cent throughout the period, with a peak during the 2012\u0026ndash;2019 time frame.\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eWhether this electoral transformation carries consequences for economic institutions and economic freedom has been a question of both academic and policy relevance (Funke, Schularick and Trebesch \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Economic freedom\u0026mdash;as measured by the Economic Freedom of the World (EFW) index of the Fraser Institute\u0026mdash;captures the degree to which individuals and enterprises operate in an conducive environment, which protects property rights, and secure sound monetary policy, limited and non-distorting government, open international trade, and limited regulation. Cross-country evidence consistently associates higher EFW scores with better economic outcomes, including higher income, faster growth, reduced poverty, and stronger investment (Lawson \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Berggren \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; de Haan and Sturm \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). The question of whether the populist wave threatens these institutional arrangements is therefore of considerable importance.\u003c/p\u003e \u003cp\u003eThe theoretical priors are ambiguous, as populism, regardless of ideological host, is primarily grounded on a conception of \u0026ldquo;the will of the people\u0026rdquo; that is inherently hostile to institutions of counter-majoritarian constraints, such as judicial independence, central bank autonomy, regulatory independence, which are themselves the underpinning of economic freedom (Mudde and Rovira Kaltwasser \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Levitsky and Ziblatt \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, the expected direction of institutional change depends critically on whether populists are in government (and can directly dictate public policy) or merely in opposition with rising vote shares (in which case the primary mechanism is signaling-driven uncertainty and reputational effects on institutions). This paper\u0026rsquo;s empirical design captures the vote-share channel, which includes both the anticipatory effects of populist electoral strength and the marginal influence that large opposition blocs exercise on incumbents\u0026rsquo; policy calculus. The distinction between these channels is elaborated in Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e2.4\u003c/span\u003e and is important for interpreting the magnitude of the estimated associations.\u003c/p\u003e \u003cp\u003eThe questions is not new in institutional economics, as there are two closely related recent studies bracketing this paper\u0026rsquo;s contribution. Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) use a global panel of electoral democracies from 1970 to 2019 and find that aggregate populism in government is significantly negatively associated with EFW changes for non-OECD countries, while in OECD settings, with strong institutional capacity, the association is largely mediated by political constraints and government ideology. Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), working in the complementary reverse direction, find that in European democracies weaker rule of law (Area 2 of EFW) is significantly associated with higher right-wing populism vote shares, however changes in aggregate economic freedom do not predict changes in populism vote shares. This paper sits at the intersection of these two studies, as elaborated in Section \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e2.3\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe results confirm that aggregate populism vote share has no statistically significant association with overall economic freedom changes in European democracies, which are institutionally more developed compared to o global panel dataset, a finding consistent with what Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) suggest for OECD-type settings. When decomposed by ideology, radical right populism vote share exhibits a significant negative direct association (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) attenuated by stronger judicial constraints. Among sub-components, Freedom to Trade Internationally shows the strongest negative association with radical right populism\u0026mdash;marginally significant at the 10 per cent level under the paper\u0026rsquo;s preferred wild cluster bootstrap inference (WCB p\u0026thinsp;=\u0026thinsp;0.063; conventional clustered p\u0026thinsp;=\u0026thinsp;0.041 with EU membership control; conventional p\u0026thinsp;=\u0026thinsp;0.070 without). Results are robust to alternative specifications.\u003c/p\u003e \u003cp\u003eI organise the rest of the paper in the following way. Section \u003cspan refid=\"Sec2\" class=\"InternalRef\"\u003e2\u003c/span\u003e reviews the relevant literature. Section \u003cspan refid=\"Sec7\" class=\"InternalRef\"\u003e3\u003c/span\u003e describes the data and variables. Section \u003cspan refid=\"Sec13\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents the empirical strategy. Section \u003cspan refid=\"Sec20\" class=\"InternalRef\"\u003e5\u003c/span\u003e reports results. Section \u003cspan refid=\"Sec24\" class=\"InternalRef\"\u003e6\u003c/span\u003e discusses findings and limitations. Section \u003cspan refid=\"Sec29\" class=\"InternalRef\"\u003e7\u003c/span\u003e concludes.\u003c/p\u003e"},{"header":"2. Literature Review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 The Concept of Economic Freedom and Its Determinants\u003c/h2\u003e \u003cp\u003eThe EFW index aggregates 42 components across five areas: (1) size of government, (2) legal system and property rights, (3) sound money, (4) freedom to trade internationally, and (5) regulation, while a substantial body of empirical research shows that higher EFW scores are robustly associated with higher per capita income, faster long-run growth, lower inflation, and better institutions (Gwartney, Lawson and Holcombe \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; de Haan and Sturm \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Hall and Lawson \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Lawson \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The determinants of EFW changes over time have received less systematic attention, though studies have identified democratic institutions (Bj\u0026oslash;rnskov and Rode \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), government ideology (Bj\u0026oslash;rnskov and Potrafke \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), income and development level (Rode and Gwartney \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), and trade openness as relevant correlates. Bj\u0026oslash;rnskov and Potrafke (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2012\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) establish an important baseline: governments of the political right are systematically associated with higher subsequent EFW scores, while left-wing governments are associated with reductions (i.e. less economic freedom), particularly in government size and regulation components. This ideological baseline is important for this study because the relationship between populism and economic freedom cannot be disentangled from the ideological character of the populist movement in question.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Populism: Conceptualisation and Measurement\u003c/h2\u003e \u003cp\u003eThe concept of populism has been operationalised through several approaches. Binary classifications (Rooduijn et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Hein\u0026ouml; \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) offer broad coverage but they fall short on information on the extent and the intensity of populist phenomenon. Celico, Rode and Rodr\u0026iacute;guez Carre\u0026ntilde;o (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) use machine-learning random forest methods to extend expert-survey data to 1,920 parties across 163 countries from 1970 to 2019, yielding a continuous 0\u0026ndash;10 scale capturing the degree of populist discourse. This study uses Timbro\u0026rsquo;s Authoritarian Populism Index (TAP, Hein\u0026ouml; \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) dataset, which provides continuous vote share data decomposed into radical right and radical left components across European democracies. This operationalisation captures electoral support rather than parliamentary representation, which allows a direct side-by-side comparison of right versus left populism effects without relying on a scalar ideology moderator. Crucially, electoral strength need not translate into government participation to generate institutional consequences. The reason is that a large populist bloc in opposition is able to constrain incumbent policy space, shift the median voter calculus, and create anticipatory uncertainty among investors and trading partners\u0026mdash;all mechanisms through which rising vote shares can affect economic freedom indicators well before any formal transfer of executive power. TAP vote share data are broadly consistent with the PopuList coding (Rooduijn et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u0026mdash;both draw on similar expert classifications of party families.\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eA related definitional caveat is the relationship between radical populism and Euroscepticism. A significant share of parties coded as radical right in TAP are simultaneously Eurosceptic (opposing deeper European integration, opposing the single market, or seeking treaty revision). To the extent that Eurosceptic rhetoric shapes trade-policy expectations\u0026mdash;over and above the protectionist motivation captured by the Area 4 channel\u0026mdash;the radical right vote share variable partially proxies Eurosceptic sentiment, to a great extent. Disentangling the protectionist from the purely sovereignty-based Eurosceptic component would require a finer-grained party-ideology measure than is available in the TAP dataset. Therefore, I flag this as a scope limitation, as the results of Freedom to Trade Internationally area may reflect a combination of protectionist policy signalling and Eurosceptic-driven uncertainty about the future trade architecture of the country.\u003c/p\u003e \u003cp\u003eAn important caveat concerns within-category heterogeneity. Parties classified as \u0026lsquo;radical right\u0026rsquo; span a wide range of economic programmes: Marine Le Pen\u0026rsquo;s Rassemblement National has historically combined welfare chauvinism with economic interventionism, while Viktor Orb\u0026aacute;n\u0026rsquo;s Fidesz has pursued a more orthodox fiscal consolidation alongside selective protectionism and state-directed industrial policy. Similarly, radical left parties extend to a wide range, from moderate social democrats to more genuinely post-capitalist formations. The radical right and radical left vote share variables therefore aggregate over heterogeneous programmes, and the estimated coefficients capture mean effects within each ideological family. This is a limitation that party-level or policy-specific analysis would help address.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Populism, Institutions, and Economic Freedom\u003c/h2\u003e \u003cp\u003eThe empirical study of populism\u0026rsquo;s economic consequences is a relatively recent enterprise. Funke, Schularick and Trebesch (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) trace 51 populist episodes since 1900 and find that countries under populist governance suffer substantially larger long-run economic costs\u0026mdash;in terms of GDP per capita, trade openness, and institutional quality\u0026mdash;with costs compounding over successive terms. More specifically about economic freedom, Rode and Revuelta (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) provide the first systematic cross-national study, finding that populism in government is negatively associated with EFW across a sample weighted toward left-populist cases in Latin American. Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) revisit this question with a more comprehensive sample and more sophisticated populism measure, and their key finding is that the association between populism and economic freedom (proxied by EFW) is significant and negative for the full democratic sample and for non-OECD countries, but in OECD countries the negative association is largely mediated by political constraints and government ideology. They interpret this as reflecting institutional safeguards, since in countries with stronger checks and balances, populist governments find it harder to erode economic freedom. Related evidence comes from St\u0026ouml;ckl and Rode (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), who find that negative financial market effects of populist elections are primarily driven by left-wing populism. Bennett et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) document that populist governments are associated with institutional erosion, corruption, and weaker property rights, with political constraints limiting these effects. This pattern is consistent with the concept of executive aggrandizement\u0026mdash;the incremental, legally cloaked erosion of counter-majoritarian institutions by elected executives who retain formal democratic legitimacy while systematically dismantling its substance (Bermeo \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2016\u003c/span\u003e)\u0026mdash;rather than with outright coups or extra-constitutional seizures of power.\u003c/p\u003e \u003cp\u003eAn important complementary strand examines the exact reverse direction. Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), in the Handbook of Research on Economic Freedom, address whether economic freedom predicts populism for 33 European democracies over 1980\u0026ndash;2020. Their panel fixed-effects analysis reveals one highly robust finding: legal system and property rights (Area 2 of the EFW index, i.e. rule of law) is significantly and negatively correlated with right-wing populism vote shares, while no other EFW area manged to achieve significance. For left-wing populism, no EFW component is significantly associated with vote shares. Examining changes over time, they find no correlation between changes in aggregate economic freedom and populist vote shares. This null result provides partial reassurance\u0026mdash;discussed in Section \u003cspan refid=\"Sec28\" class=\"InternalRef\"\u003e6.4\u003c/span\u003e\u0026mdash;that the reverse causality channel in this specification is not sufficiently operative.\u003c/p\u003e \u003cp\u003eThe two studies\u0026mdash;Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u0026mdash;frame the contribution of the present paper, as it differs from both along three distinct dimensions. First, on sample scope and direction of inquiry: unlike Celico and Rode\u0026rsquo;s global panel of 2,367 observations, this paper focuses specifically on European democracies, which serve as a) the institutional setting directly relevant to the contemporary populist wave and b) as the set of countries with the most developed check and balances mechanisms in western democracies\u0026mdash;while using a longer time horizon (to 2023) than Bergh and K\u0026auml;rn\u0026auml; (to 2020); and, like Celico and Rode, it models populism as a predictor of EFW changes, complementing rather than replicating Bergh and K\u0026auml;rn\u0026auml;\u0026rsquo;s reverse-direction framework. Second, on the treatment of ideology: both prior papers use a single scalar measure or separate single-equation estimates for right versus left, whereas this study includes radical right and radical left vote shares simultaneously in the same equation. Third, and most distinctively, I conduct sub-component regressions for each of the five EFW areas as separate outcomes\u0026mdash;asking not only whether aggregate economic freedom is affected but which specific institutional channels are identified as the most sensitive to populist electoral pressure. This sub-component analysis is, to the best of author\u0026rsquo;s knowledge, absent from earlier research and from the broader comparative literature on populism and economic institutions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Theoretical Mechanisms: Policy Channel and Signalling Channel\u003c/h2\u003e \u003cp\u003eA theoretical concern with vote-share specifications requires explicit acknowledgment, since the effect of populist electoral strength on economic freedom may operate through two conceptually distinct mechanisms, which the current design captures jointly rather than separately. The first is the direct policy channel: populists who enter the parliament and cabinets, can affect policy making by legislate policies\u0026mdash;protectionist tariffs, central bank subordination, property right abridgements, expanded transfers\u0026mdash;that mechanically alter the EFW sub-scores directly. The second is a signalling channel: rising populist vote shares create anticipatory uncertainty among investors, trading partners, and institutional actors even before populists reach office, generating reputational and expectation effects that can affect economic freedom indicators. Such effects could be the widening of sovereign spreads, reduced foreign direct investment, or voluntary regulatory forbearance by incumbents seeking to neutralise the populist threat.\u003c/p\u003e \u003cp\u003eThe empirical design\u0026mdash;using continuous vote shares (not government participation) as the key regressor\u0026mdash;is best interpreted as capturing the combined electoral-strength effect, including both the direct policy effects of populists in government and the signalling effects of their presence as a large electoral force. I include a robustness check using a binary Populists in Government indicator (Section \u003cspan refid=\"Sec17\" class=\"InternalRef\"\u003e4.4\u003c/span\u003e), which isolates the direct governance channel. The Discussion interprets results in light of both mechanisms, treating them as one. The public choice literature offers a further theoretical mechanism for the trade freedom channel specifically: protectionism as coalition-building, where radical right parties favour domestically organised producers over diffuse consumer interests\u0026mdash;a mechanism that can operate both through direct policy (tariff legislation) and through electoral signalling to incumbent governments facing populist competition (Autor et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Guriev and Papaioannou \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Data and Variables","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Dependent Variable: Economic Freedom of the World\u003c/h2\u003e \u003cp\u003eThe primary dependent variable is the annual change in the Economic Freedom of the World (EFW) index (Gwartney, Lawson, Hall and Murphy \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), available annually from 2000 onwards on a 0\u0026ndash;10 scale. I model the change ΔEFW\u0026thinsp;=\u0026thinsp;EFW_t\u0026thinsp;\u0026minus;\u0026thinsp;EFW_{t-1} as the dependent variable rather than the level, following Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and Rode and Revuelta (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). For sub-component analysis, I use ΔEFW Areas 1\u0026ndash;5 as separate outcomes. Raw controls for sub-component regressions are drawn from the very EFW 2025 Master Index Dataset: government consumption as a share of GDP (raw component of Area 1, entered as its natural logarithm given the right-skewed distribution); the annual CPI inflation rate (raw component of Area 3, entered as log(1\u0026thinsp;+\u0026thinsp;π/100) \u0026times; 100 to accommodate the 62 country-year observations with negative inflation values while compressing the high-inflation tail); and the mean applied tariff rate (raw component of Area 4, entered in percentage levels). The tariff rate is not log-transformed because several country-year observations record zero values; a log(1\u0026thinsp;+\u0026thinsp;tariff) transformation was tested in robustness and does not end up with significant different results.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Populism Variables\u003c/h2\u003e \u003cp\u003ePopulism data come from the Timbro\u0026rsquo;s Authoritarian Populism index (TAP, Hein\u0026ouml; \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) dataset covering European democracies. I use: (1) overall populism vote share (POP_VOTE), the combined vote share of all populist parties; (2) radical right vote share (RRIGHT_VOTE); and (3) radical left vote share (RLEFT_VOTE). Vote shares are recorded at election years and carried forward until the next election. In the sample, mean overall populism vote share is 19.3 per cent (SD\u0026thinsp;=\u0026thinsp;15.2 pp, as reported in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), rising from approximately 12 per cent in 2001 to 25 per cent by 2023, with the increase primarily driven by radical right electoral support. Radical right parties average 12.1 per cent (SD\u0026thinsp;=\u0026thinsp;13.4 pp); radical left parties average 6.6 per cent (SD\u0026thinsp;=\u0026thinsp;9.5 pp). Populists participate in government in approximately 21 per cent of country-year observations.\u003c/p\u003e \u003cp\u003eTAP records vote shares for parties that meet the dataset\u0026rsquo;s populism-classification threshold; for country-years where no qualifying radical right or radical left party was present, I assign a vote share of zero, as the most methodologically appropriate coding when the absence of a qualifying party implies zero populist electoral mass. Under this zero-fill convention the full (N\u0026thinsp;=\u0026thinsp;713) and sub-component (N\u0026thinsp;=\u0026thinsp;713) analytic samples are identical: all 31 \u0026times; 23 country-year cells contain complete populism vote share data. A covariate balance check confirms that the two samples are statistically indistinguishable across all key covariates (see Appendix Table A1). The note that TAP is broadly consistent with the PopuList coding (Rooduijn et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) extends to Eurosceptic parties as well: parties coded as radical right or radical left in TAP typically appear under closely related headings in PopuList\u0026rsquo;s Eurosceptic category, though the two databases draw different party-family boundaries and a formal overlap matrix has not been computed for this paper.\u003c/p\u003e \u003cp\u003eI also need to clarify that the carry-forward convention warrants explicit theoretical justification. What this specification measures is electoral strength as an institutional signal: the 33 per cent vote share of the Rassemblement National in France in 2022 does not cease to \u0026lsquo;press\u0026rsquo; on the policy calculus of the Macron government in 2023 and 2024\u0026mdash;incumbents adjust continuously in anticipation of a large opposition bloc, not only at election years. This is precisely the signalling channel outlined in Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e2.4\u003c/span\u003e. The limitation of the approach is that it does not capture within-term polling fluctuations, which would constitute a higher-frequency measure of populist pressure.\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Institutional Controls\u003c/h2\u003e \u003cp\u003eI use the V-Dem Judicial Constraints index (Coppedge et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) as the primary institutional moderator. The Judicial Constraints (JC) index measures the extent to which courts constrain executive authority, ranging from 0 to 1. I think this measure fits better because it is theoretically more precise for this paper\u0026rsquo;s question, which concerns the mechanism through which institutional constraints moderate the relationship between populism and EFW, operates primarily through the judiciary\u0026rsquo;s capacity to prevent executives from dismantling property rights and monetary independence. In the sample, the mean Judicial Constraints index is 0.903 (SD\u0026thinsp;=\u0026thinsp;0.082), with considerable within-country variation over time, particularly in countries experiencing democratic backsliding. Critically, the interaction between populism and judicial constraints is primarily identified from countries with meaningful JC variation over the sample period\u0026mdash;most prominently Hungary and Poland, where the JC index declined significantly after 2010 (Hungary) and 2015 (Poland) as the respective governments moved against judicial independence.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Additional Controls\u003c/h2\u003e \u003cp\u003eI include the lagged EFW level (or sub-area level) and the lagged natural logarithm of GDP per capita in PPP terms (2021 international dollars) as standard controls. In addition, I include the lagged EU membership dummy (1\u0026thinsp;=\u0026thinsp;EU member in year t\u0026thinsp;\u0026minus;\u0026thinsp;1, 0 otherwise), which takes value 1 continuously for longstanding members and transitions from 0 to 1 at accession for the 14 countries that joined the EU during the panel period (including Bulgaria and Romania in 2007, Croatia in 2013, and the 2004 cohort). Although the EU dummy is largely absorbed by country fixed effects for longstanding members, it captures accession-related policy transitions for the 14 countries with within-sample variation. For Sound Money (Area 3), I note that EU membership is an imperfect proxy for monetary regime: Eurozone members face ECB-disciplined inflation, whereas non-Eurozone EU members retain independent monetary policy. The EU dummy thus conflates Eurozone and non-Eurozone membership for Area 3 analysis, and a more granular monetary-regime indicator would be advisable in future extended work. The EU dummy is not significant in any main specification.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Sample\u003c/h2\u003e \u003cp\u003eThe intersection of the TAP dataset with the EFW panel yields a sample of 31 European democracies from 2001 to 2023, producing up to 712 country-year observations after lagging (Models 1\u0026ndash;2) or 663 observations (Models 3\u0026ndash;4, where the right/left decomposition is available). Twenty-six of the 31 countries are OECD members. This homogeneous sample of institutionally mature high-income European democracies allows to focus on a demanding test for the populism\u0026ndash;economic freedom nexus: if effects are present here, they persist despite strong institutional safeguards. Importantly, the homogeneity of the sample also implies attenuation bias: coefficients estimated on a sample with limited variance in both the JC index (confined to a narrow high range) and the EFW outcome (compressed by institutional stability) are likely downward-biased in absolute value relative to what would be found in a more diverse global panel. Namely, this test can be treated as a conservative estimation for our results. Descriptive statistics are in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive Statistics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003e Variable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMin\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMax\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ EFW Overall (dep. var.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.835\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.850\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ Area 1: Government Size\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.246\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;1.320\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.977\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ Area 2: Legal System \u0026amp; Property Rights\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.353\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.855\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ Area 3: Sound Money\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.447\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;2.456\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.263\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ Area 4: Freedom to Trade Internationally\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.228\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;1.236\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.471\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eΔ Area 5: Regulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.211\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;1.697\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.717\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEFW Overall (level)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7.623\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.424\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5.753\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.670\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulism Vote Share (%, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e19.281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15.211\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e69.500\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Right Vote Share (%, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12.132\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.396\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e69.400\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Left Vote Share (%, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6.598\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.483\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e45.100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulists in Government (0/1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.208\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEU Membership (0/1, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.776\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.417\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJudicial Constraints Index (V-Dem, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.903\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.507\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.988\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLiberal Component Index (V-Dem, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.905\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.658\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.984\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLog GDP per capita PPP (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10.701\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.427\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e9.439\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e11.840\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLog Govt Consumption (% GDP, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3.269\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.219\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.084\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.652\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCPI Inflation Rate (%, t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.284\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;4.480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e34.468\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote: Sample covers 31 European democracies, 2001\u0026ndash;2023. Δ denotes annual first difference. All lagged variables from t\u0026thinsp;\u0026minus;\u0026thinsp;1. Sources: Fraser Institute EFW 2025; TAP; V-Dem v16; World Bank WDI.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Empirical Strategy","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Identification Approach\u003c/h2\u003e \u003cp\u003eThe empirical design exploits within-country, within-year variation after removing country-specific time-invariant characteristics and common time trends. Concretely, the two-way fixed effects (TWFE) estimator demeans each variable by subtracting the country mean, the year mean, and adding back the overall mean (the within transformation). The remaining variation\u0026mdash;\u0026lsquo;within-country variation relative to the cross-country and cross-year patterns\u0026rsquo;\u0026mdash;is what identifies the coefficients. In practical terms: the radical right coefficient is identified by years within a given country when its radical right vote share was above or below that country\u0026rsquo;s own average, controlling for years when all countries experienced common shocks. Countries that have not experienced meaningful variation in populism vote share over the period contribute little identifying power; the variation is primarily supplied by countries such as Hungary, Italy, France, Sweden, and Greece, whose radical right shares have moved substantially since 2001.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.2 The Baseline Specification\u003c/h2\u003e \u003cp\u003eThe baseline estimating equation follows the change specification of Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and Rode and Revuelta (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), augmented with EU membership:\u003c/p\u003e \u003cp\u003e \u003cem\u003eΔEFW_it\u0026thinsp;=\u0026thinsp;β₁ EFW_{t-1} + β₂ POP_{t-1} + β₃ JC_{t-1} + β₄ logGDPpc_{t-1} + β₅ EU_{t-1} + δᴵ + γₜ + ε_it\u003c/em\u003e \u003c/p\u003e \u003cp\u003ewhere ΔEFW_it is the annual change in the EFW overall index; EFW_{t-1} is the lagged EFW level capturing mean-reversion; POP_{t-1} is the lagged populism vote share; JC_{t-1} is the lagged Judicial Constraints index; logGDPpc_{t-1} is the lagged log GDP per capita; EU_{t-1} is the lagged EU membership indicator; and δᴵ, γₜ are country and year fixed effects respectively. Standard errors are clustered at the country level.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Ideology Decomposition and Interactions\u003c/h2\u003e \u003cp\u003eIn Model 4, I replace aggregate POP with radical right (RRIGHT) and radical left (RLEFT) vote shares and add ideology-by-JC interactions:\u003c/p\u003e \u003cp\u003e \u003cem\u003eΔEFW\u0026thinsp;=\u0026thinsp;β₁EFW_{t-1} + β₂RRIGHT_{t-1} + β₃RLEFT_{t-1} + β₄(RRIGHT\u0026times;JC)_{t-1} + β₅(RLEFT\u0026times;JC)_{t-1} + β₆JC_{t-1} + β₇logGDPpc_{t-1} + β₈EU_{t-1} + δᴵ + γₜ + ε\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe marginal effect of radical right populism is \u0026part;(ΔEFW)/\u0026part;RRIGHT\u0026thinsp;=\u0026thinsp;β₂ + β₄ \u0026times; JC. A positive β₄ suggests that stronger judicial constraints moderate (attenuate) the negative association. Note that β₂ is the coefficient when JC\u0026thinsp;=\u0026thinsp;0, a hypothetical value far outside the observed sample range (minimum observed JC\u0026thinsp;=\u0026thinsp;0.507); it should not be interpreted as the effect at any empirically plausible institutional level. The estimated marginal effects at observed JC quantiles are reported in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Sub-component Regressions\u003c/h2\u003e \u003cp\u003eI run five sub-component regressions replacing ΔEFW Overall with ΔEFW Area k (k\u0026thinsp;=\u0026thinsp;1,...,5). Area-specific controls are included: log government consumption (% GDP) for Area 1, size of goverment; log(1\u0026thinsp;+\u0026thinsp;CPI inflation/100) \u0026times; 100 for Area 3, sound money; mean applied tariff rate for Area 4, freedom to trade. Areas 2 (legal system and property rights) and 5 (regulation) do not include area-specific raw controls because the principal sub-components of those areas (legal institution ratings, regulatory quality scores) are themselves the outcomes of interest rather than inputs available as separate regressors in the EFW master dataset. All specifications include EU membership and country and year fixed effects.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Robustness\u003c/h2\u003e \u003cp\u003eThree robustness checks are reported: (1) three-year rolling average ΔEFW as dependent variable to smooth annual EFW noise\u0026mdash;I note that this creates a temporal mismatch since the populism regressor remains annual; a cleaner test would match three-year rolling averages of both variables, which is proposed for future work; (2) binary Populists in Government indicator instead of continuous vote share, which isolates the direct governance channel; (3) V-Dem Liberal Component Index (LCI) substituted for Judicial Constraints, as a more generic mechanism of institutional constraints, than include parliament and independents authorities.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.6 Estimation, Inference, and Caveats\u003c/h2\u003e \u003cp\u003eModels are estimated using PanelOLS from the Python linearmodels 7.0 package with within-transformation for entity and time effects. Standard errors are clustered at the country level. With 31 clusters, conventional clustered SEs fall below the threshold of approximately 50 clusters recommended for reliable inference (Cameron, Gelbach and Miller \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; MacKinnon \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Wild cluster bootstrap (WCB) SEs with 4,999 Rademacher draws (Roodman et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) are therefore implemented and reported in Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e as the preferred inference baseline. The Wooldridge (2002) test for first-order serial correlation in panel residuals yields rho\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.040 (H₀: rho\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.5; t\u0026thinsp;=\u0026thinsp;9.38, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), confirming the presence of AR(1) serial correlation. Country-clustered standard errors, which permit arbitrary within-country serial correlation without restricting the autocorrelation structure, are the appropriate response to this finding; the WCB provides additional robustness under the Rademacher distribution. The Arellano-Bond system-GMM estimator, which additionally addresses Nickell (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e1981\u003c/span\u003e) bias and the endogeneity of the populism regressor, remains the recommended extension for future work with a larger cross-section. One feature of the WCB results warrants clarification: the WCB standard error for the Model 1 radical right coefficient (0.00252) is marginally smaller than the conventional clustered SE (0.00260). This is not a reporting error, because with 31 clusters, the conventional asymptotic finite-sample correction can slightly over-inflate the variance estimate, while the Rademacher-draw bootstrap distribution, which is computed directly from within-cluster residuals, converges on a tighter empirical spread under certain configurations of within-cluster serial correlation. The WCB p-value (0.030) is arithmetically consistent with the marginally smaller bootstrap SE.\u003c/p\u003e \u003cp\u003eI must note that despite the fact that several statistical tests have been applied to ensure as robust relationships as possible, the results should be interpreted as associations, since no causal model was employed. Two main endogeneity concerns apply: (1) reverse causality\u0026mdash;declining EFW attracting populist voters\u0026mdash;and (2) omitted time-varying confounders. As discussed in Section \u003cspan refid=\"Sec28\" class=\"InternalRef\"\u003e6.4\u003c/span\u003e, a panel Granger-causality test is implemented in this version. Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) find no robust panel evidence that EFW changes predict populism vote share changes in European democracies; the extended Granger tests in Section \u003cspan refid=\"Sec28\" class=\"InternalRef\"\u003e6.4\u003c/span\u003e corroborate this finding within the sample across multiple lag structures.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Results","content":"\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Main Models: Overall EFW\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents four nested models. Three patterns emerge consistently. First, the lagged EFW level is negative and highly significant (coefficient\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.25, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), confirming mean-reversion. Second, the Judicial Constraints index is positive and significant in Models 1 and 3 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.10), indicating that stronger judicial oversight correlates with subsequent economic freedom improvements. Third, log GDP per capita and EU membership are statistically insignificant throughout, reflecting limited within-country variation once country fixed effects are absorbed.\u003c/p\u003e \u003cp\u003eAggregate populism vote share in Model 1 has a coefficient of \u0026minus;\u0026thinsp;0.00038 (SE\u0026thinsp;=\u0026thinsp;0.00051, p\u0026thinsp;=\u0026thinsp;0.462)\u0026mdash;small, negative, and indistinguishable from zero. This confirms the finding of Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) for OECD-type settings that aggregate populism does not significantly correlate with EFW changes in high-income institutionally constrained democracies. In Model 2, the interaction of populism vote share with Judicial Constraints is similarly negligible. In Model 3, both radical right (\u0026minus;\u0026thinsp;0.00076, p\u0026thinsp;=\u0026thinsp;0.310) and radical left (\u0026minus;\u0026thinsp;0.00008, p\u0026thinsp;=\u0026thinsp;0.922) vote shares are individually insignificant.\u003c/p\u003e \u003cp\u003eModel 4, the most informative specification, includes ideology-by-JC interaction terms. The direct coefficient on radical right vote share is \u0026minus;\u0026thinsp;0.00444 (SE\u0026thinsp;=\u0026thinsp;0.00206, p\u0026thinsp;=\u0026thinsp;0.031 conventional; WCB p\u0026thinsp;=\u0026thinsp;0.030, Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e)\u0026mdash;statistically significant at the 5 per cent level under both conventional and bootstrap inference. The interaction RRIGHT \u0026times; JC is positive and marginally significant (0.00419, SE\u0026thinsp;=\u0026thinsp;0.00227, p\u0026thinsp;=\u0026thinsp;0.065). The marginal effect of radical right populism is therefore: \u0026part;(ΔEFW)/\u0026part;RRIGHT\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.00444\u0026thinsp;+\u0026thinsp;0.00419 \u0026times; JC. At the sample mean (JC\u0026thinsp;=\u0026thinsp;0.902), the marginal effect is \u0026minus;\u0026thinsp;0.00060 per percentage point of vote share; multiplied by one standard deviation (13.5 pp), this implies\u0026thinsp;\u0026minus;\u0026thinsp;0.008 EFW points per year, or approximately\u0026thinsp;\u0026minus;\u0026thinsp;0.03 EFW points over a four-year electoral cycle.\u003c/p\u003e \u003cp\u003eThe radical left coefficient in Model 4 is positive and not significant (0.01764, p\u0026thinsp;=\u0026thinsp;0.124). For the radical left null result, I note that the observed mean radical left vote share is 6.6 per cent (SD\u0026thinsp;=\u0026thinsp;9.5 pp)\u0026mdash;substantially lower prevalence than radical right in this European sample. Under the Model 4 specification, a one-standard-deviation increase in radical left vote share (9.5 pp) would imply an effect of only 0.00167 EFW points per year at the sample mean JC\u0026mdash;an economically negligible magnitude even if statistically non-zero. The null result thus reflects both a baseline-shift mechanism (discussed in Section \u003cspan refid=\"Sec26\" class=\"InternalRef\"\u003e6.2\u003c/span\u003e) and insufficient statistical power given lower prevalence and variance of radical left vote shares. One caveat merits acknowledgment: the Rad. Left \u0026times; JC interaction term, while not significant under conventional clustered inference (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), yields WCB p\u0026thinsp;=\u0026thinsp;0.054 (Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e)\u0026mdash;borderline at the 10 per cent level. This does not alter the substantive interpretation\u0026mdash;the radical left coefficient on ΔEFW remains economically negligible at all empirically plausible JC values, however, the marginal bootstrap result warrants cautious acknowledgment rather than a conclusive null finding. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e visualises the marginal effects of the main finding, that of the radical right populism on ΔEFW at selected JC levels.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMain Regression Results: Annual Change in EFW Overall\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eM1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eM2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eM3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eM4\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEFW (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.2491***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.2493***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.2559***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.2559***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0302)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0313)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0316)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0321)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulism Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.0004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0005)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0035)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePop \u0026times; Judicial Constraints (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0039)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Right Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0044**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0007)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Left Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0176\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0009)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0115)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Right \u0026times; Judicial Constraints (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0042*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0023)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Left \u0026times; Judicial Constraints (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0203\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0135)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLog GDP per capita (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.0188\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.0188\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0017\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0520)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0520)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0509)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0551)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJudicial Constraints (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.1436**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.1498\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.0931*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0150\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0630)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.1752)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0537)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0989)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEU Membership (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0096\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0097\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.0068\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0096\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0225)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0225)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0213)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0225)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWithin-R\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.254\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.254\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.233\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.237\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry FE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYear FE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClusters\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e\u003cem\u003eNote: Dependent variable: annual change in EFW overall index. All regressors lagged one period. Country-clustered standard errors in parentheses. *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; * p\u0026thinsp;\u0026lt;\u0026thinsp;0.10. PanelOLS with entity and time fixed effects. M3\u0026ndash;4 use 663 obs. due to availability of radical right/left decomposition in TAP. The M3 aggregate null for radical right and left is consistent with the M1 result, suggesting sample attrition does not drive findings.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMarginal Effects of Radical Right Populism on ΔEFW (Model 4)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJC Percentile / Value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eJC Value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMarginal Effect (per pp)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1 SD Effect (\u0026times;13.5 pp)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10th percentile (institutionally weaker)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.792\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.00107\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0144\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25th percentile\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.863\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.00074\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSample mean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.902\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.00060\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0082\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e75th percentile\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.957\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.00036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0049\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e90th percentile (institutionally stronger)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.978\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.00029\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0039\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCrossover (ME\u0026thinsp;=\u0026thinsp;0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.047 (out of sample)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eNote: Marginal effect = \u0026part;(ΔEFW)/\u0026part;(Rad. Right Vote Share)\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.00444\u0026thinsp;+\u0026thinsp;0.00419 \u0026times; JC (Model 4). \u0026ldquo;1 SD Effect\u0026rdquo; is the change in ΔEFW associated with a one SD increase in radical right vote share (13.5 pp). The crossover at JC\u0026thinsp;=\u0026thinsp;1.047 lies outside the observed sample range (max\u0026thinsp;=\u0026thinsp;0.988), confirming the negative association is present at all empirically plausible JC levels.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Sub-component Regressions\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e presents sub-component regressions using aggregate populism vote share. No sub-component shows a statistically significant association with aggregate populism vote share at conventional levels. Coefficients are uniformly small, ranging from \u0026minus;\u0026thinsp;0.00252 (Area 3: Sound Money, p\u0026thinsp;=\u0026thinsp;0.175) to +\u0026thinsp;0.00080 (Area 5: Regulation, p\u0026thinsp;=\u0026thinsp;0.485). In Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the Judicial Constraints index is significant for Areas 2 and 3 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 each), confirming that stronger judicial oversight is associated with improvements in both the legal system and sound money sub-components, independent of populism. The EU membership variable is not significant for any sub-component, though the directional coefficient for Area 3 is positive (0.109, p\u0026thinsp;=\u0026thinsp;0.187).\u003c/p\u003e \u003cp\u003eIn the ideology decomposition (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), Area 4 (Freedom to Trade Internationally) exhibits a negative association with radical right populism vote share (coefficient\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.00251, SE\u0026thinsp;=\u0026thinsp;0.00119, conventional p\u0026thinsp;=\u0026thinsp;0.041). Under the preferred wild cluster bootstrap inference (Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), the WCB p-value is 0.063\u0026mdash;marginally significant at the 10 per cent level but not crossing the conventional 5 per cent threshold. The conventional p-value without the EU membership control is 0.070. WCB p\u0026thinsp;=\u0026thinsp;0.063 should be treated as the definitive estimate for this coefficient; the conventional p\u0026thinsp;=\u0026thinsp;0.041 is reported for comparability with specifications that do not implement bootstrap inference. Area 1 (Government Size) shows a negative but insignificant coefficient for radical left populism (\u0026minus;\u0026thinsp;0.00330, p\u0026thinsp;=\u0026thinsp;0.247), consistent in direction with left-populist expansion of government spending, however, the power of this relationship is insufficient for a firm conclusion. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e visualises the radical right and radical left populism coefficients across EFW sub-components (Table \u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e specification).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSub-component Regressions: Aggregate Populism Vote Share\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea 1 Gov. Size\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArea 2 Legal/Prop.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea 3 Snd. Money\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eArea 4 Trade\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eArea 5 Regulation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEFW Area (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.188***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.308***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.135***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.206***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;0.182***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.039)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.045)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.032)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.035)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.036)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulism Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.0019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.0008\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0007)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0009)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.0011)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJudicial Constraints (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.249**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.420**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.147)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.116)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.177)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.108)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.100)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEU Membership (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.109\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.048\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.040)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.084)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.054)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.031)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLog GDP p.c. (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.055\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.053\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.150)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.062)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.109)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.067)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.073)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArea-specific control\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLog govt cons.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLog(1\u0026thinsp;+\u0026thinsp;infl)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e712\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWithin-R\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.061\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.099\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.052\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry \u0026amp; Year FE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote: Dependent variables: annual changes in EFW sub-areas 1\u0026ndash;5. All regressors lagged one period. Country-clustered SEs in parentheses. *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; * p\u0026thinsp;\u0026lt;\u0026thinsp;0.10.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSub-component Regressions: Ideology Decomposition\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea 1 Gov. Size\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArea 2 Legal/Prop.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea 3 Snd. Money\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eArea 4 Trade\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eArea 5 Regulation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Right Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.0009\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.0003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0025**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.0009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0022)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0009)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0022)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0012)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.0010)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadical Left Vote Share (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.0033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0010\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.0005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.0015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.0029)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.0009)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.0041)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.0010)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.0028)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEFW Area (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.188***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.289***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;0.131***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;0.202***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;0.172***\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.040)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.046)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.032)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.036)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.038)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEU Membership (t\u0026thinsp;\u0026minus;\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.025\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.046\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.040)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.022)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.085)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.055)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.031)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e663\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWithin-R\u0026sup2;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.097\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.095\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.051\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry \u0026amp; Year FE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote: Dependent variables: annual changes in EFW sub-areas 1\u0026ndash;5. Radical right and left vote shares included simultaneously. Country-clustered SEs in parentheses. *** p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; * p\u0026thinsp;\u0026lt;\u0026thinsp;0.10. The Area 4 radical right coefficient (p\u0026thinsp;=\u0026thinsp;0.041) is borderline: p\u0026thinsp;=\u0026thinsp;0.070 without EU membership control. JC coefficient suppressed for brevity; full results available upon request.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRobustness Checks\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpecification\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoeff. on Populism\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline (M1): Annual ΔEFW, vote share, EU control\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00051\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.462\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRobustness 1: 3-yr rolling avg ΔEFW, vote share\u0026sup1;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00054\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.689\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRobustness 2: Annual ΔEFW, binary populist-in-gov\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00932\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.377\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRobustness 3: LCI instead of JC as institutional proxy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00026\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00051\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.610\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eNote: All models include country and year FEs, EU membership dummy, clustered SEs (31 clusters). \u0026sup1;Rolling average creates temporal mismatch (annual populism vs. smoothed EFW); a matched rolling-average populism variable would constitute a cleaner test.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eWild Cluster Bootstrap Inference (4,999 Rademacher Draws)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable / Model\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoeff.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eClustered SE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eClust. p\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWCB SE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eWCB p\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM1: Rad. Right Vote Share\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00484\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00260\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00252\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.030**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM1: Rad. Left Vote Share\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00384\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00227\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.091\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.054+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM4: Rad. Right \u0026times; JC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.01827\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01040\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.080\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00988\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.065+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM4: Rad. Left Vote Share\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.03409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.02315\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.141\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.02201\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.130\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM4: Rad. Left \u0026times; JC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.04349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.02472\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.079\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.02351\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.054+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArea 4: Rad. Right Vote Share\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00609\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00360\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.093\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00343\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.063+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArea 4: Rad. Left Vote Share\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00592\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.147\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.152\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote: Wild cluster bootstrap via double-demeaned within-transformation with Rademacher cluster-sign draws (B\u0026thinsp;=\u0026thinsp;4,999, seed\u0026thinsp;=\u0026thinsp;42). WCB p-values are two-sided. ** WCB p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; + WCB p\u0026thinsp;\u0026lt;\u0026thinsp;0.10.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e5.3 Backsliding versus Stable-Democracy Heterogeneity\u003c/h2\u003e \u003cp\u003eSection \u003cspan refid=\"Sec14\" class=\"InternalRef\"\u003e4.1\u003c/span\u003e notes that the JC\u0026times;RRIGHT interaction in Model 4 is primarily identified from countries experiencing democratic backsliding\u0026mdash; that is Hungary (JC decline post-2010) and Poland (JC decline post-2015). Hungary\u0026rsquo;s V-Dem Judicial Constraints index fell from approximately 0.92 in 2010 to 0.78 by 2020, a decline of roughly 14 percentage points on the 0\u0026ndash;1 scale; Poland\u0026rsquo;s index declined from approximately 0.90 in 2015 to below 0.80 by 2021. Both countries are placed well below the sample mean of 0.903 by the end of the backsliding period, providing the primary within-country variation that identifies the backsliding interaction. To assess whether the radical right\u0026ndash;EFW relationship differs systematically between backsliding and stable democracies, I augment the Model 1 specification with a backsliding dummy (BS\u0026thinsp;=\u0026thinsp;1 for Hungary from 2010, Poland from 2015, and 0 otherwise) and its interactions with radical right and radical left vote shares. The results are reported in Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe base radical right coefficient (effect in stable democracies) is \u0026minus;\u0026thinsp;0.00519 (SE\u0026thinsp;=\u0026thinsp;0.00306, p\u0026thinsp;=\u0026thinsp;0.090), directionally consistent with Model 1 and marginally significant. The interaction RRIGHT\u0026times;BS is positive and statistically significant at the 5 per cent level (0.00649, SE\u0026thinsp;=\u0026thinsp;0.00269, p\u0026thinsp;=\u0026thinsp;0.016), indicating that the negative association between radical right vote share and EFW changes is substantially attenuated during backsliding episodes. The combined marginal effect during backsliding periods is approximately zero (+\u0026thinsp;0.00130). This pattern is consistent with the constitutional capture mechanism: as Orb\u0026aacute;n\u0026rsquo;s Fidesz and Kaczyński\u0026rsquo;s PiS consolidated power, they selectively liberalised business regulation and pursued supply-side tax reductions\u0026mdash;policies that improve EFW Area 1 and Area 5 scores\u0026mdash;while simultaneously eroding judicial independence and trade openness. The near-zero net effect during backsliding reflects this mixed institutional legacy: democratic regression (negative for Area 2) partly offset by business-friendly economic deregulation (positive for Areas 1 and 5). The radical left interaction RLEFT\u0026times;BS is negative but imprecisely estimated (\u0026minus;\u0026thinsp;0.155, SE\u0026thinsp;=\u0026thinsp;0.093, p\u0026thinsp;=\u0026thinsp;0.097).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBacksliding Heterogeneity: Radical Right vs. Stable Democracy Split\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoeff.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Right Vote Share (stable democracies)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00306\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.090+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Left Vote Share (stable democracies)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.00393\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00227\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.084+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBacksliding Dummy (HUN 2010+, POL 2015+)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00650\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.042**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Right \u0026times; Backsliding\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00649\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00269\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.016**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad. Left \u0026times; Backsliding\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.15469\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.09297\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.097+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eME of Rad. Right during backsliding\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00130\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026mdash;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObservations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCountry \u0026amp; Year FE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eNote: Dependent variable: annual ΔEFW overall. Backsliding dummy\u0026thinsp;=\u0026thinsp;1 for Hungary\u0026thinsp;\u0026ge;\u0026thinsp;2010 and Poland\u0026thinsp;\u0026ge;\u0026thinsp;2015, 0 otherwise. ME during backsliding\u0026thinsp;=\u0026thinsp;base RR coeff\u0026thinsp;+\u0026thinsp;RR\u0026times;BS coeff\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.00519\u0026thinsp;+\u0026thinsp;0.00649\u0026thinsp;=\u0026thinsp;+\u0026thinsp;0.00130. Country-clustered SEs. ** p\u0026thinsp;\u0026lt;\u0026thinsp;0.05; + p\u0026thinsp;\u0026lt;\u0026thinsp;0.10.\u003c/em\u003e\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"6. Discussion","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\n\u003ch2\u003e6.1 The European Panel as a Demanding Test\u003c/h2\u003e\n\u003cp\u003eThe results contribute to the literature in a specific and bounded way. The null result for aggregate populism in European democracies can be interpreted as precisely what was expected by the theoretical framework and what Celico and Rode (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e) find for OECD settings. European democracies in the sample have high and relatively stable Judicial Constraints scores (mean JC\u0026thinsp;=\u0026thinsp;0.903), and the institutional safeguards present in most of these polities appear sufficient to absorb or deflect, at least, short-run populist-driven policy changes. The homogeneity of the sample, which is a design strength by choice for internal validity, simultaneously suggests attenuation bias: the true coefficient on radical right populism in a more institutionally diverse global sample is likely larger in absolute value than the conservative estimates presented here.\u003c/p\u003e\n\u003cp\u003eI note the significance of the crossover JC value (1.047, outside the sample range) with more nuance than a simple report might convey. Its implication is not that judicial constraints are irrelevant, they demonstrably attenuate the negative association as the interaction coefficient shows, but rather that no European democracy in the sample sits at a JC level high enough to fully neutralise the negative association of radical right populism. At the 90th percentile of JC (0.978), the 1 SD annual effect is only\u0026thinsp;\u0026minus;\u0026thinsp;0.004 EFW points: statistically non-zero at current coefficient estimates but economically negligible. The appropriate conclusion is therefore: stronger judicial institutions systematically attenuate the association, but even the institutionally strongest European democracies do not eliminate it entirely.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec26\" class=\"Section2\"\u003e\n\u003ch2\u003e6.2 Trade Freedom as the Most Susceptible Sub-component\u003c/h2\u003e\n\u003cp\u003eThe significant negative association between radical right populism and Area 4 (Trade Freedom, WCB p\u0026thinsp;=\u0026thinsp;0.063; conventional p\u0026thinsp;=\u0026thinsp;0.041 with EU membership control, p\u0026thinsp;=\u0026thinsp;0.070 without) is the strongest sub-component finding. This result requires a theoretical reconciliation with Bergh and K\u0026auml;rn\u0026auml;\u0026rsquo;s (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) emphasis on Area 2 (rule of law) as the central nexus between economic freedom and right-wing populism. The two findings are not contradictory but operate in different directions and timeframes: Bergh and K\u0026auml;rn\u0026auml; find that countries with weaker rule of law produce more right-wing populism (the reverse direction, using EFW levels), while the results here show that rising radical right vote shares are associated with deteriorations in trade freedom (the direction of this paper, using EFW changes). Trade Freedom is the sub-component most directly actionable through electoral and governmental channels: tariff rates, non-tariff barriers, and regulatory trade restrictions are standard legislative instruments that governments facing populist pressure may deploy in the short run, without necessarily requiring the constitutional or judicial modifications needed to erode Area 2.\u003c/p\u003e\n\u003cp\u003eThe public choice mechanism can be instructive here: protectionism serves as a coalition-building instrument for radical right parties, directing concentrated benefits to politically organised domestic industries (manufacturing, agriculture) while dispersing costs over unorganised consumers and trading partners. This mechanism operates both through direct policy when populists govern and through incumbents\u0026rsquo; anticipatory concessions to electoral pressure\u0026mdash;explaining why the vote-share specification captures a significant signal even absent full government participation. The null Area 2 result in this specification does not contradict Bergh and K\u0026auml;rn\u0026auml;; it reflects the short-run nature of this annual change specification and the difficulty of capturing gradual institutional erosion in a first-differenced framework.\u003c/p\u003e\n\u003cp\u003eThe empirical literature provides strong direct support for this protectionist characterisation of European radical right parties. Colantone and Stanig (\u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e) show that regional exposure to Chinese import competition across 15 Western European countries significantly increased vote shares for nationalist and right-wing populist parties\u0026mdash;and that these parties subsequently adopted explicitly isolationist and protectionist platforms in response to their electorate\u0026rsquo;s trade-displaced support base. Dippel, Gold, Heblich and Pinto (\u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e) replicate this mechanism for Germany, tracing the path from import exposure to far-right votes via xenophobic attitudes. Rodrik (\u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e) provides the unifying theoretical account: European right-wing populism is fundamentally a form of economic nationalism, combining hostility to trade openness with demands for national industrial protection, and operating through the specific channel of tariffs, non-tariff barriers, and resistance to multilateral trade agreements\u0026mdash;precisely the components measured by EFW Area 4.\u003c/p\u003e\n\u003cp\u003eThe null finding for radical left populism in Area 4 is best understood through a baseline-shift argument rather than simply as a statistical failure. Left parties and their mainstream centre-left counterparts were already substantially converged on anti-trade, pro-intervention platforms well before the populist surge of the 2010s: the incumbent centre-left had been absorbing redistributive, trade-sceptical, and capital-control preferences into its programme for decades, although a part of the party family during 1990s (the third way) saw liberalization as a positive policy choice. A surge in radical left vote share therefore creates little measurable policy distance\u0026mdash;the counterfactual starting point is already close to the populist position. The radical right, by contrast, represented a genuine ideological rupture from a centre-right tradition that had been broadly pro-trade and pro-globalisation since the Thatcher-CDU-Maastricht consensus of the 1980s and 1990s (Inglehart and Norris \u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e; Rodrik \u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e). The shift of the right coalition toward protectionism was discontinuous and electorally driven, generating measurable policy distance and thus a detectable Area 4 signal in the data. This is consistent with the party competition literature on incumbent responsiveness to flanking parties: incumbents adjust platforms in proportion to the ideological distance they must travel (Adams et al., \u003cspan class=\"CitationRef\"\u003e2004\u003c/span\u003e; Somer-Topcu \u003cspan class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eA secondary mechanism reinforces the left null: the most prominent left-populist governments in the sample, that is SYRIZA-ANEL in Greece (2015\u0026ndash;2019) and the Podemos-supported coalition in Spain, governed under EU fiscal surveillance, Eurozone membership, and, in the Greek case, full Troika conditionality. These external constraints effectively compressed the policy space available to left populists in power, forcing a degree of programme moderation that their electoral mandate did not reflect. SYRIZA under Tsipras ultimately signed austerity measures that were in complete opposite direction of its electoral programme a few months earlier; similarly, Podemos in coalition was constrained by EU deficit rules. This \u0026lsquo;moderation in office\u0026rsquo; dynamic is structurally asymmetric: Orb\u0026aacute;n\u0026rsquo;s Hungary and Kaczyński\u0026rsquo;s Poland were net EU budget recipients facing no equivalent external fiscal disciplining mechanism, and Article 7 proceedings against them required unanimity\u0026mdash;a far weaker constraint than Troika conditionality. The right populist programmes could therefore be partially implemented, while the left populist programmes largely could not. This asymmetry in external constraint, rather than any inherent moderation of left populism, accounts in part for the differential signals in the data.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec27\" class=\"Section2\"\u003e\n\u003ch2\u003e6.3 Comparison with Celico and Rode (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e) and Bergh and K\u0026auml;rn\u0026auml; (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/h2\u003e\n\u003cp\u003eThis study was primarily initiated to discuss with the results of the tow prior studies. The findings align with Celico and Rode (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e) on the aggregate null for OECD democracies, while seeking to provide a more granular decomposition on both ideology and EFW components. Their finding that the association between populism and EFW in OECD settings is mediated by political ideology is corroborated by the ideology decomposition: the null aggregate result in Models 1\u0026ndash;3 masks a conditional negative direct effect of radical right populism that is moderated by judicial constraints (Model 4). The main difference is that the results show a consistently null result for left-wing populism in the European panel\u0026mdash;likely because left-populist episodes (SYRIZA, Podemos) are less prevalent and shorter-lived in the sample than the historically left-dominant populism that drove Rode and Revuelta\u0026rsquo;s (\u003cspan class=\"CitationRef\"\u003e2015\u003c/span\u003e) results, and because of the external constraint asymmetry described in Section \u003cspan class=\"InternalRef\"\u003e6.2\u003c/span\u003e. As discussed in Section \u003cspan class=\"InternalRef\"\u003e6.2\u003c/span\u003e above, While Bergh and K\u0026auml;rn\u0026auml; (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) identify weak rule of law as the structural condition breeding right-wing populism, this study finds that once populists gain electoral ground, it is trade freedom \u0026mdash; the most legislatively accessible dimension of economic freedom \u0026mdash; that deteriorates first.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec28\" class=\"Section2\"\u003e\n\u003ch2\u003e6.4 Limitations\u003c/h2\u003e\n\u003cp\u003eSeveral limitations must be acknowledged. First, all results reflect correlational relationships with no causal effects mechanisms, by design. The potential for reverse causality, that is declining economic freedom contributing to populist electoral success, cannot be fully excluded by lagged regressors alone. Although Bergh and K\u0026auml;rn\u0026auml; (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) find no robust panel evidence that changes in aggregate economic freedom predict changes in populism vote shares across European democracies, I additionally implement a Dumitrescu\u0026ndash;Hurlin (2012) panel Granger-causality test\u0026mdash;asking whether lags of radical right (or radical left) vote share predict EFW changes, country-by-country, with the average F-statistic aggregated into a W-statistic\u0026mdash;across three lag structures (2, 3, and 5 years) to assess robustness at short to medium-run horizons. Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003e reports the results. Across all lag structures and for both vote-share variables, the test consistently fails to reject H₀ of no panel Granger causality (all p\u0026thinsp;\u0026gt;\u0026thinsp;0.90). These results strongly corroborate the signalling-channel interpretation that populist vote shares do not lead EFW changes, at least, in the short to medium run; any causal pathway would require a longer horizon than five-year lags can detect, since huge institutional changes require a long term erosion; maybe Hungary\u0026rsquo;s 16 years case could provide such a institutional ground for further research. In addition, a credible structural causal identification strategy\u0026mdash;instrumental variables or regression discontinuity around close populist election victories\u0026mdash;remains an important agenda for future work.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\u0026nbsp;\u003c/div\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab10\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eExtended Dumitrescu\u0026ndash;Hurlin Panel Granger Causality Tests\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eLags\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eVariable\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eavg-F\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eW\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ep-value\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eG (countries)\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eN (obs)\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Right Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.445\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e44.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.951\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e620\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Left Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.890\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e26.7\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e\u0026gt;\u0026thinsp;0.99\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e30\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e600\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Right Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.271\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e39.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e\u0026gt;\u0026thinsp;0.99\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e589\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Left Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.800\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e24.8\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e\u0026gt;\u0026thinsp;0.99\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e589\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Right Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.659\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e51.4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e\u0026gt;\u0026thinsp;0.99\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e527\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eRadical Left Vote Share\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.021\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31.6\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e\u0026gt;\u0026thinsp;0.99\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e31\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e527\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003ctfoot\u003e\n\u003ctr\u003e\n\u003ctd colspan=\"7\"\u003e\u003cem\u003eNote: Dumitrescu\u0026ndash;Hurlin (2012) panel Granger test. H₀: populism vote shares do not Granger-cause \u0026Delta;EFW. W\u0026thinsp;=\u0026thinsp;sum of country-level F-statistics; p-value based on chi\u0026sup2;(G\u0026times;lags) approximation. G\u0026thinsp;=\u0026thinsp;number of countries with sufficient observations. All p-values well above 0.10 across all lag structures, confirming the absence of a short-to-medium-run reverse-causality channel.\u003c/em\u003e\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tfoot\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eSecond, the sample of 31 European countries is geographically restricted, by selection design. Findings should not be generalised to Latin American, African, or Asian democracies where the institutional and ideological character of populism differs substantially. Third, with 31 country clusters, conventional clustered standard errors may be insufficiently conservative; wild cluster bootstrap inference is implemented in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e and generally confirms the directional results, with the radical right coefficient in Model 1 strengthening to WCB p\u0026thinsp;=\u0026thinsp;0.030. Fourth, the within-term constancy of populism vote share\u0026mdash;carried forward from election years\u0026mdash;limits annual within-term variation; the robustness checks suggest this does not drive the results. Fifth, the Nickell (\u003cspan class=\"CitationRef\"\u003e1981\u003c/span\u003e) bias from the lagged dependent variable is expected to be negligible at T\u0026thinsp;\u0026asymp;\u0026thinsp;23 years, but an Arellano-Bond GMM estimator remains the recommended extension. Sixth, non-electoral channels of populist influence\u0026mdash;referendum campaigns (Brexit), executive decrees, and supra-parliamentary mobilisation\u0026mdash;are not captured by vote-share or government-participation indicators. Seventh, the JC\u0026times;RRIGHT interaction is primarily identified from countries experiencing democratic backsliding (Hungary, Poland); Section \u003cspan class=\"InternalRef\"\u003e5.3\u003c/span\u003e presents the backsliding heterogeneity analysis, finding that the negative radical right\u0026ndash;EFW association is present in stable democracies (p\u0026thinsp;=\u0026thinsp;0.090) and attenuated in backsliding episodes. Eighth, the Wooldridge (2002) test confirms first-order serial correlation in panel residuals (rho\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.040 vs. H₀: rho\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.5; t\u0026thinsp;=\u0026thinsp;9.38, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Country-clustered standard errors and the wild cluster bootstrap address this.\u003c/p\u003e\n\u003cp\u003eOn the question of structural identification an ideal instrument for radical right party strength would require variation in the supply of populist candidates or party funding that is exogenous to both domestic economic conditions and EFW trajectories. The strategic foreign funding of intellectual and political infrastructure by sympathetic foreign governments\u0026mdash;channelled through foundations and non-governmental organizations indirectly, rather than direct party finance, represents one possible identification structure; nevertheless, the endogenous targeting of already-receptive environments and the opacity of financial flows would likely undermine the exclusion restriction in practice.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"7. Conclusion","content":"\u003cp\u003eThis paper has examined the association between populism and economic freedom across 31 European democracies from 2001 to 2023, extending the framework of Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and complementing the reverse-direction analysis of Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). I introduced three novel dimensions: simultaneous ideology decomposition (radical right vs. radical left), sub-component regressions for all five EFW areas, and a backsliding heterogeneity analysis distinguishing institutionally stable democracies from democratic backsliders. EU membership was included as an additional time-varying control, while the signalling versus policy channel distinction was made explicit in the theoretical framework.\u003c/p\u003e \u003cp\u003eAggregate populism vote share is not significantly associated with annual EFW changes, confirming the institutional mediation prediction for OECD-type democracies with high institutional framework. When decomposed by ideology, radical right populism has a significant direct negative association attenuated by judicial constraint strength\u0026mdash;economically small at the sample mean but potentially cumulative over electoral cycles. The sub-component analysis identifies Freedom to Trade Internationally as the EFW component most sensitive to radical right populism (marginally significant at the 10 per cent level under WCB inference: WCB p\u0026thinsp;=\u0026thinsp;0.063; conventional p\u0026thinsp;=\u0026thinsp;0.041), consistent with the protectionist orientation of contemporary European radical right platforms and the public choice logic of protectionism as coalition-building. The radical left null result is consistent with a baseline-shift mechanism\u0026mdash;left parties were already converged on anti-trade positions, leaving no measurable policy distance to cover\u0026mdash;reinforced by the external constraints that disciplined left-populist governments in practice (Troika conditionality, Eurozone fiscal rules), an asymmetry not shared by the right populist governments in Hungary and Poland.\u003c/p\u003e \u003cp\u003eThe paper contributes a focused European test, an explicit side-by-side comparison of the two most closely related predecessor studies, and the first sub-component analysis in this literature to identify which EFW channels are most sensitive to populist electoral strength. Future research priorities should employ (1) Arellano-Bond GMM robustness and stronger causal identification through electoral quasi-experiments; (2) a global panel extension; (3) longer-horizon effects on EFW legal system and sound money areas, which theory suggests are more vulnerable to sustained populist governance than to short-run electoral pressure; (4) party-level analysis to disaggregate within-family heterogeneity in economic programmes; and (5) a matched rolling-average specification pairing populism and EFW changes at the same temporal frequency.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe author declares no competing interests, financial or non-financial, related to the work submitted for publication.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConstantinos Saravakos conceived and designed the study, collected and processed the data, conducted all empirical analyses, interpreted the results, and wrote the manuscript in its entirety.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe author thanks Robert Lawson (Jerome M. Fullinwider Centennial Chair in Economic Freedom and Director of the Bridwell Institute for Economic Freedom, Cox School of Business, Southern Methodist University), Dr. Dan Mitchell (Center for Freedom and Prosperity), and Ioannis Filios for helpful comments and suggestions on earlier drafts of this manuscript. The usual disclaimer applies.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data that support the findings of this study are drawn from four publicly available sources. Economic freedom data are from the Fraser Institute Economic Freedom of the World 2025 Annual Report and Master Index Dataset, available at https://www.fraserinstitute.org/economic-freedom/dataset. Populism vote share data are from Timbro Authoritarian Populism Index the https://populismindex.com/. Institutional data are from the Varieties of Democracy (V-Dem) Dataset v16, available at https://www.v-dem.net/data/the-v-dem-dataset. Macroeconomic controls are from the World Bank World Development Indicators, available at https://databank.worldbank.org/source/world-development-indicators. Python code for the wild cluster bootstrap and Dumitrescu\u0026ndash;Hurlin panel Granger tests was developed with AI-assisted technical support (Claude, Anthropic).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbsher S, Grier K, Grier R (2020) The economic consequences of durable left-populist regimes in Latin America. J Econ Behav Organ 177:787\u0026ndash;817\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdams J, Clark M, Ezrow L, Glasgow G (2004) Understanding change and stability in party ideologies: Do parties respond to public opinion or to past election results? Br J Polit Sci 34(4):589\u0026ndash;610\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAutor D, Dorn D, Hanson G, Majlesi K (2020) Importing political polarization? The electoral consequences of rising trade exposure. Am Econ Rev 110(10):3139\u0026ndash;3183\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBennett DL, Boudreaux C, Nikolaev B (2021) Populist discourse and entrepreneurship: The role of political ideology and institutions. J Int Bus Stud 52(4):800\u0026ndash;810\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBennett DL, Boudreaux C, Nikolaev B (2023) Populism and economic freedom. J Comp Econ 51(2):574\u0026ndash;597\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerggren N (1999) Economic freedom and equality: friends or foes? Public Choice 100(3):203\u0026ndash;223\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBergh A, K\u0026auml;rn\u0026auml; A (2024) Economic freedom and populism. In: Berggren N (ed) Handbook of Research on Economic Freedom. Edward Elgar, Cheltenham, pp 297\u0026ndash;314\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBermeo N (2016) On democratic backsliding. J Democr 27(1):5\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBj\u0026oslash;rnskov C, Potrafke N (2012) Political ideology and economic freedom across Canadian provinces. East Econ J 38(2):143\u0026ndash;166\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBj\u0026oslash;rnskov C, Potrafke N (2013) The size and scope of government in the US states: does party ideology matter? Int Tax Public Finance 20(4):687\u0026ndash;714\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBj\u0026oslash;rnskov C, Rode M (2020) Regime types and regime change: A new dataset on democracy, coups, and political institutions. Rev Int Organ 15(2):531\u0026ndash;551\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCameron AC, Gelbach JB, Miller DL (2008) Bootstrap-based improvements for inference with clustered errors. Rev Econ Stat 90(3):414\u0026ndash;427\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCelico A, Rode M (2023) Populism, majority rule, and economic freedom. In: Gwartney J, Lawson R, Murphy R (eds) Economic Freedom of the World: 2023 Annual Report. Fraser Institute, Vancouver\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCelico A, Rode M, Rodr\u0026iacute;guez Carre\u0026ntilde;o J (2022) Random forest populism indicators. Working Paper, Universidad de Navarra\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColantone I, Stanig P (2018) The trade origins of economic nationalism: Import competition and voting behavior in Western Europe. Am J Polit Sci 62(4):936\u0026ndash;953\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoppedge M et al (2023) V-Dem [Country-Year/Country-Date] Dataset v13. Varieties of Democracy (V-Dem) Project\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Haan J, Sturm J-E (2000) On the relationship between economic freedom and economic growth. Eur J Polit Econ 16(2):215\u0026ndash;241\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDippel C, Gold R, Heblich S, Pinto R (2019) The path from trade to right-wing populism in Europe. Elect Stud 60:102040\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDumitrescu E-I, Hurlin C (2012) Testing for Granger non-causality in heterogeneous panels. Econ Model 29(4):1450\u0026ndash;1460\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFunke M, Schularick M, Trebesch C (2023) Populist leaders and the economy. Am Econ Rev 113(12):3249\u0026ndash;3288\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuriev S, Papaioannou E (2022) The political economy of populism. J Econ Lit 60(3):753\u0026ndash;832\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGwartney J, Lawson R, Hall J, Murphy R (2023) Economic Freedom of the World: 2023 Annual Report. Fraser Institute, Vancouver\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGwartney J, Lawson R, Holcombe R (1999) Economic freedom and the environment for economic growth. J Inst Theor Econ 155(4):643\u0026ndash;663\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall JC, Lawson RA (2014) Economic freedom of the world: An accounting of the literature. Contemp Econ Policy 32(1):1\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalikiopoulou D, Vlandas T (2019) What is new and what is nationalist about Europe\u0026rsquo;s new nationalism? Explaining the rise of the far right in Europe. Nations Natl 25(2):409\u0026ndash;434\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHein\u0026ouml; AJ (2024) Authoritarian Populism Index 2024. Timbro, Stockholm. www.populismindex.com\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenisz WJ (2000) The institutional environment for economic growth. Econ Polit 12(1):1\u0026ndash;31\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuber RA, Schimpf CH (2017) On the distinct effects of left-wing and right-wing populism on democratic quality. Polit Gov 5(4):146\u0026ndash;165\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInglehart R, Norris P (2016) Trump, Brexit, and the rise of populism: Economic have-nots and cultural backlash. HKS Working Paper RWP16-026. Harvard Kennedy School, Cambridge\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJ\u0026auml;ger K (2017) Economic freedom in the early 21st century: Government ideology still matters. Kyklos 70(2):257\u0026ndash;278\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson R (2022) Economic freedom of the world: Annual update. Indep Rev 27(1):3\u0026ndash;12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLevitsky S, Ziblatt D (2018) How Democracies Die. Crown, New York\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026uuml;hrmann A et al (2020) Varieties of Party Identity and Organization (V-Party) Dataset V1. Varieties of Democracy. V-Dem) Project\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacKinnon JG (2019) How cluster-robust inference is changing applied econometrics. Can J Econ 52(3):851\u0026ndash;881\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMargalit Y (2019) Economic insecurity and the causes of populism, reconsidered. J Econ Perspect 33(4):152\u0026ndash;170\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeijers MJ, Zaslove A (2021) Measuring populism in political parties: Appraisal of a new approach. Comp Polit Stud 54(2):372\u0026ndash;407\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMudde C, Rovira Kaltwasser C (2017) Populism: A Very Short Introduction. Oxford University Press, Oxford\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNickell S (1981) Biases in dynamic models with fixed effects. Econometrica 49(6):1417\u0026ndash;1426\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePitlik H (2007) A race to liberalization? Diffusion of economic policy reform among OECD-economies. Public Choice 132(1):159\u0026ndash;178\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRode M, Gwartney J (2012) Does liberalization cause growth? An empirical investigation using EFW data. J Priv Enterp 27(2):1\u0026ndash;28\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRode M, Revuelta J (2015) The wild bunch! An empirical note on populism and economic institutions. Econ Gov 16(1):73\u0026ndash;96\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodrik D (2018) Populism and the economics of globalization. J Int Bus Policy 1(1\u0026ndash;2):12\u0026ndash;33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoodman D, Nielsen M\u0026Oslash;, MacKinnon JG, Webb MD (2019) Fast and wild: Bootstrap inference in Stata using boottest. Stata J 19(1):4\u0026ndash;60\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRooduijn M et al (2019) The PopuList: An overview of populist, far right, far left and Eurosceptic parties in Europe. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003c/span\u003e\u003cspan address=\"http://www.popu-list.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSomer-Topcu Z (2009) Timely decisions: The effects of past national elections on party policy change. J Polit 71(1):238\u0026ndash;248\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSt\u0026ouml;ckl T, Rode M (2021) All that glitters: Populism, political risk, and the IMF. Kyklos 74(1):50\u0026ndash;76\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTeam Populism (2024) TAP Dataset: Populist, Radical Right and Radical Left Parties in European Democracies. Brigham Young University, Provo\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeyland K (1999) Neoliberal populism in Latin America and Eastern Europe. Comp Polit 31(4):379\u0026ndash;401\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Footnotes","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eThroughout this paper, I employ the term \u0026lsquo;far-right\u0026rsquo; as a definition for these political formations, following Halikiopoulou and Vlandas (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), who demonstrate that \u0026lsquo;far-right\u0026rsquo; captures both the populist and the nationalist-authoritarian dimensions that are jointly operative in parties such as Fidesz, PiS, and Lega. In tables and variable labels I retain \u0026lsquo;radical right\u0026rsquo; to match the TAP 2024 dataset coding, which uses that term for the relevant party family.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e However, I note that no formal cross-dataset correlation exercise is conducted here.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e A harmonised polling series for 31 countries over 23 years is not currently available; the robustness checks (binary government-participation indicator, three-year rolling EFW average) provide partial validation that the carry-forward convention does not drive the results.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"populism, economic freedom, radical right, panel data, democratic backsliding","lastPublishedDoi":"10.21203/rs.3.rs-9451071/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9451071/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper seeks to examine the relationship between populism and economic freedom across 31 European democracies, during 2001\u0026ndash;2023, extending the research by Celico and Rode (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and complementing Bergh and K\u0026auml;rn\u0026auml; (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The paper contributes three novel dimensions: simultaneous ideological decomposition (radical right vs. radical left); sub-component regressions for all five EFW areas; and a backsliding heterogeneity split (Hungary from 2010, Poland from 2015). Aggregate populism vote share shows no significant association with annual EFW changes, however, when decomposed by ideological host, radical right populism exhibits a significant negative direct association (WCB p\u0026thinsp;=\u0026thinsp;0.030) attenuated by stronger judicial constraints, concentrated in stable democracies and near-zero during backsliding episodes (rr \u0026times; backsliding interaction p\u0026thinsp;=\u0026thinsp;0.016). This finding is consistent with the mixed institutional legacy of Orb\u0026aacute;n- and PiS-style governance. Among sub-components, Freedom to Trade Internationally shows the strongest negative association with radical right populism (WCB p\u0026thinsp;=\u0026thinsp;0.063), consistent with the protectionist orientation of contemporary European far-right platforms. The null result for radical left populism may reflect a baseline-shift mechanism: mainstream and radical left parties were already converged on anti-trade positions, leaving no measurable policy distance, asymmetrically reinforced external conditionality that disciplined left-populist parties in ways that had no equivalent for right-populist incumbents. A Dumitrescu\u0026ndash;Hurlin panel Granger test at 2, 3, and 5 lags consistently fails to reject the null of no Granger causality from populist vote shares to EFW changes, an indication of strong correlational associations, but not a causal effects identification.\u003c/p\u003e","manuscriptTitle":"Populism and Economic Freedom in European Democracies: Evidence from a Two-Way Fixed Effects Panel Analysis, 2001–2023","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-27 13:59:37","doi":"10.21203/rs.3.rs-9451071/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"209d75ca-e305-408c-8470-51bd6ff6957c","owner":[],"postedDate":"April 27th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-09T06:55:11+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-09T07:11:23+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-27 13:59:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9451071","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9451071","identity":"rs-9451071","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}