| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1052264 | Electoral Studies | 2011 | 17 Pages |
In the last 20 years, extreme right parties have gained a substantial share of electorate in a number of Western European countries. Despite of substantial research interest, the reasons for the popularity of the Extreme Right parties presented in the studies are often contradictory. Our study has a two-fold contribution to the topic. First, we show that studies based on aggregate data analysis may yield biased results that are not confirmed when contextual and individual-level data are fitted together using the growth curve analysis. Second, we argue that the analysis of extreme-right vote based on one election suggests a constant effect of contextual variables on the outcome variable. We demonstrate that the effects of local context vary over time, and hence should be modeled using the multilevel (growth curve) approach.
► We model extreme-right vote using ecological and cross-sectional time-series data over time. ► These data are modeled using Tobit and multilevel (growth curve) analysis. ► We discuss and compare the outcomes of both models. ► Ecological analysis overestimates the role of local context and suggests its effects are static. ► Multilevel analysis suggests the effects are dynamic and outcomes are time-specific
