class: center, middle, inverse, title-slide # Deliberation Across the World ## Examining the Link Between
Regime Legitimacy and Deliberation
Rosa Seitz
1
& Fabio Votta
2
23-08-2018
###
1
E-Mail:
rosa.marie.seitz@gmail.com
2
E-Mail:
fabio.votta@gmail.com
###
Slides:
https://deliberation-and-legitimacy.netlify.com/
--- <style> .onehundredtwenty { font-size: 120%; } <style> .ninety { font-size: 90%; } .eightyfive { font-size: 85%; } .eighty { font-size: 80%; } .seventyfive { font-size: 75%; } .seventy { font-size: 70%; } </style> <br> <br> <br> <br> <br> <br> <br> <br> .onehundredtwenty[ Research Question: > What role does the deliberative quality of a political system play for regime legitimacy? ] --- class: inverse, center, middle ## Theoretical Link between <br> Deliberation and Regime Legitimacy --- ### Theory > Deliberation in its original sense contains the process of ***rational consideration of arguments***, in which participants provide ***reasoned justification*** for their positions, relate to the ***common good***, ***respect other opinions*** and are willing to ***yield to better arguments*** .center[ <i> for an overview see Bächtiger et al. (2010) </i> ] -- + Central claim of deliberative theory: to arrive at (more) legitimate decisions .center[ <i> see Habermas (1994), Chambers (1996) & Parkinson (2006) </i> ] -- + Legitimacy conceptualized sociologically: regime support -- + General assumption: *deliberative qualities* of a political system increase support from its citizens -- + Consequences of *Authoritarian Deliberation* (He & Warren, 2011) + Authoritarian regimes might use deliberation to *increase* support + Might *decrease* support due to increasing democratic capabilites --- class: inverse, center, middle ## Data & Methodology --- #### Data & Methodology **Individual-Level Data** + Merging of the following datasets: + Afrobarometer Survey, Round 5 and Round 6 (data from 2011-2015) + Asian Barometer Survey, Wave 3 and Wave 4 (2010-2015) + AmericasBarometer (2010-2014) + European Social Survey, Round 6 (2012) + Latinobarometro (2015) + World Values Survey (2010-2014) + In total 316,938 individual cases in 119 Countries + weighted to same sample size (=1000) -- **Country-Level Data** + "Varieties of Democracy"-Dataset: Deliberative Component Index + Control variables come from V-Dem and the QoG-Dataset --- ## Operationalization of Regime Support <br> **Factor Analysis:**
Variable
Loadings
Trust in Political Leadership
0.72
Trust in Police
0.76
Trust in Courts
0.64
Trust in Parliament
0.75
Variance explained: 52%
Cronbach's α: 0.81
--- ## Deliberative Component Index (DCI) <center> <img src="images/dci_table.png"> </center> --- #### Data & Methodology .ninetyfive[ + Control Variables + Micro Level + Age + Sex (Male/Female) + Financial Security + Education + Employment (0/1) + Macro Level + Polity/FH + Democracy (0/1)<sup>1</sup> + Autocracy (0/1)<sup>1</sup> + logged GDP per capita + logged Population + Life Expectancy + Urban Pop. Ratio + Survey dummies + WVS, Afrobarometer, Latinobarometro, Americasbarometer, Asianbarometer, ESS ] <font size='1'><sup>1</sup>We apply the Polity2-classification of democracies, anocracies and autocracies. Polity/FH was rescaled to range from -10 to +10 and the cut-off values are set at 6 and -6 (-10 - -6 = Autocracy; -5 - 5 = Anocracy; 6 - 10 = Democracy).</font> --- class: inverse, center, middle ## Analytical Challenges --- ## Analytical Challenges *Two major challenges emerged:* 1. Collinearity between Deliberation indicators and Democracy Measures -- 2. Possibly biased self-reported regime support -- <br> #### 1st Challenge: Multicollinearity <img src="images/corr_table.png"> --- #### 1st Challenge: Correlation between DCI and Polity/FH <center> <img src="images/delib_polity.png"> </center> --- #### 2nd Challenge: Possibly biased self-reported Regime Support <img src="images/doubleplot.png"> --- #### 2nd Challenge: Regime Support and Discussion Suppression <img src="images/regimesupport_fod.png"> --- #### 2nd Challenge: Weighting of Regime Support <br> <img src="images/weight.png"> --- class: inverse, center, middle ## Analysis & Results --- ## Analysis Given the hierarchical nature of the data we estimate *multilevel models* + Dependent Variables + Regime Support *(No Bias)* -> `ICC: 44.97%` + Regime Support *(Low Bias)* -> `ICC: 41.70%` + Regime Support *(High Bias)* -> `ICC: 40.97%` -- + Main independent variables + Deliberative Component Index (DCI) + plus all five components of the DCI -- + In total 91 seperate models + split into three subsamples (Complete, Democracies, Non-Democracies) --- ## Results <a href="https://raw.githubusercontent.com/favstats/delib_slides/master/images/slides_coefplot.png" target = "blank"> <img src="images/slides_coefplot1.png"> </a> *Click on picture for larger Version* --- ## Results <a href="https://raw.githubusercontent.com/favstats/delib_slides/master/images/slides_coefplot.png" target = "blank"> <img src="images/slides_coefplot.png"> </a> *Click on picture for larger Version* --- ### Model Comparisons -Just Polity/FH Models vs. Complete Models <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison_polities.png" target = "blank"> <img src="images/model_comparison_polities_emp.png"> </a> *Click on picture for larger Version* --- ### Model Comparisons -Just Polity/FH Models vs. Complete Models <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison_polities.png" target = "blank"> <img src="images/model_comparison_polities_emp2.png"> </a> *Click on picture for larger Version* --- class: inverse, center, middle ## Conclusions & Future Research --- ## Conclusions + Complete Sample: + Positive effects on regime support when controlling for democracy + Otherwise small and insignificant effects + Democracy Sample: + Deliberative quality has a clear *positive effect* + Most robust results + Non-Democracy Sample: + Ambigious results (mostly insignificant) + Only *Common Good* Indicator shows some effect --- ## Future Research * Presumably biased regime support + should be taken into account + possible remedies? (e.g. weighting procedures, survey design) + individual-level vs. country-level weighting? * Validity of the DCI + more sensitive approaches to measure deliberation on the country level are neccessary + nevertheless interesting differences regarding the subsamples --- class: inverse, center, middle ## Thanks for Listening! Full Paper: <a href='https://www.academia.edu/36656020/Deliberation_Across_the_World_A_Cross-National_Examination_of_the_Link_Between_Deliberation_and_Regime_Legitimacy'>Deliberation Across the World</a> Slides: <a href='deliberation-and-legitimacy.netlify.com'>https://deliberation-and-legitimacy.netlify.com </a> Online Appendix can be found <a href='https://favstats.github.io/delib_mod_database'>here</a> Code and Data on <a href='https://github.com/favstats/paper_delib'>GitHub</a> --- ## References *Büchtiger, A., Niemeyer, S., Neblo, M., Steenbergen, M. R., & Steiner, J. (2010). Disentangling diversity in deliberative democracy: Competing theories, their blind spots and complementarities. Journal of Political Philosophy, 18(1), 32-63.* *Chambers, S. (1996). Reasonable democracy: Jürgen Habermas and the politics of discourse. Cornell University Press.* *Habermas, J. (1994). Three normative models of democracy. Constellations, 1 (1), 1-10.* *He, B., & Warren, M. E. (2011). Authoritarian deliberation: The deliberative turn in chinese political development. Perspectives on Politics, 9 (2), 269-289.* *Parkinson, J. (2006). Deliberating in the real world: Problems of legitimacy in deliberative democracy. Oxford University Press on Demand.* *Tannenberg, M. (2017). The autocratic trust bias: Politically sensitive survey items and self-censorship. Varieties of Democracy Institute: Working Paper, (49), 1-31.* --- class: inverse, center, middle ## Appendix --- #### Operationalization of Regime Support <img src="images/operat.png"> --- ### Maps - Deliberation <a href="https://github.com/favstats/delib_slides/raw/master/images/map_dci.png" target = "blank"><img src="images/map_dci.png"></a> --- ### Maps - Regime Support <a href="https://github.com/favstats/delib_slides/raw/master/images/map_regime.png" target = "blank"><img src="images/map_regime.png"></a> --- ### Coefficent Plots - Complete Sample <a href="https://github.com/favstats/delib_slides/raw/master/images/coefplot.png" target = "blank"> <img src="images/coefplot.png"></a> --- ### Coefficent Plots - Democracy Sample <a href="https://github.com/favstats/delib_slides/raw/master/images/coefplot_dem.png" target = "blank"> <img src="images/coefplot_dem.png"></a> --- ### Coefficent Plots - Non-Democracy Sample <a href="https://github.com/favstats/delib_slides/raw/master/images/coefplot_nondem.png" target = "blank"> <img src="images/coefplot_nondem.png"></a> --- ### Model Comparisons - Just Control Models vs. Just Deliberation Models <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison_wo_delibs2.png" target = "blank"> <img src="images/model_comparison_wo_delibs2.png"></a> --- ### Model Comparisons - Complete Sample <br> <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison.png" target = "blank"> <img src="images/model_comparison.png"></a> --- ### Model Comparisons - Democracy Sample <br> <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison_dem.png" target = "blank"> <img src="images/model_comparison_dem.png"></a> --- ### Model Comparisons - Non-Democracy Sample <br> <br> <a href="https://github.com/favstats/delib_slides/raw/master/images/model_comparison_nondem.png" target = "blank"> <img src="images/model_comparison_nondem.png"></a>